Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • View all journals
  • Explore content
  • About the journal
  • Publish with us
  • Sign up for alerts
  • News & Views
  • Published: 28 July 2022

FUNCTIONAL CONNECTIVITY

Traveling and standing waves in the brain

  • Javier Gonzalez-Castillo   ORCID: orcid.org/0000-0002-6520-5125 1  

Nature Neuroscience volume  25 ,  pages 980–981 ( 2022 ) Cite this article

3852 Accesses

4 Citations

6 Altmetric

Metrics details

  • Computational biology and bioinformatics
  • Neuroscience

Studying the natural wanderings of the living brain is extremely challenging. Bolt et al. describe a new framework for considering the brain’s intrinsic activity based on the geophysical concepts of standing and traveling waves.

Economists, politicians, CEOs, scientists: they all strive to understand complex dynamical systems so that they can predict — and hopefully improve — their future outcomes. They collect relevant data and morph it into actionable information using meticulously selected analytical tools. Yet, over time, conflicting views will probably emerge from the same data, owing to either unaccounted-for noise or methodological discrepancies. When this happens, it is key to promptly identify and address these issues. This is what Bolt et al. 1 set out to do in this issue of Nature Neuroscience for the field of resting-state functional MRI (fMRI), and more generally, for our understanding of intrinsic brain dynamics.

This is a preview of subscription content, access via your institution

Relevant articles

Open Access articles citing this article.

On-chip phonon-magnon reservoir for neuromorphic computing

  • Dmytro D. Yaremkevich
  • , Alexey V. Scherbakov
  •  …  Manfred Bayer

Nature Communications Open Access 14 December 2023

Access options

Access Nature and 54 other Nature Portfolio journals

Get Nature+, our best-value online-access subscription

24,99 € / 30 days

cancel any time

Subscribe to this journal

Receive 12 print issues and online access

195,33 € per year

only 16,28 € per issue

Rent or buy this article

Prices vary by article type

Prices may be subject to local taxes which are calculated during checkout

brain travelling waves

Bolt, T. et al. Nat. Neurosci. https://doi.org/10.1038/s41593-022-01118-1 (2022).

Article   PubMed   Google Scholar  

Ogawa, S. et al. Biophys. J. 64 , 803–812 (1993).

Article   CAS   Google Scholar  

Lu, H., Golay, X., Pekar, J. J. & Van Zijl, P. C. M. Magn. Reson. Med. 50 , 263–274 (2003).

Article   Google Scholar  

Caballero-Gaudes, C. & Reynolds, R. C. Neuroimage 154 , 128–149 (2017).

Biswal, B., Yetkin, F. Z., Haughton, V. M. & Hyde, J. S. Magn. Reson. Med. 34 , 537–541 (1995).

Gonzalez-Castillo, J. et al. Neuroimage 202 , 116129 (2019).

Abbas, A. et al. Neuroimage 191 , 193–204 (2019).

Liu, T. T., Nalci, A. & Falahpour, M. Neuroimage 150 , 213–229 (2017).

Liu, X., Zhang, N., Chang, C. & Duyn, J. H. Neuroimage 180 , 485–494 (2018).

Fox, M. D. et al. Proc. Natl Acad. Sci. USA 102 , 9673–9678 (2005).

Margulies, D. S. et al. Proc. Natl Acad. Sci. USA 113 , 12574–12579 (2016).

Smith, S., Miller, K., Moeller, S. & Xu, J. Proc. Natl Acad. Sci. USA 109 , 3131–3136 (2012).

Vidaurre, D., Smith, S. M. & Woolrich, M. W. Proc. Natl Acad. Sci. USA 114 , 12827–12832 (2017).

Gonzalez-Castillo, J., Kam, J. W. Y., Hoy, C. W. & Bandettini, P. A. J. Neurosci. 41 , 1130–1141 (2021).

Download references

Acknowledgements

J.G.-C. was supported by the Intramural Research Program of the National Institute of Mental Health (annual report ZIAMH002783).

Author information

Authors and affiliations.

Section on Functional Imaging Methods, National Institute of Mental Health, National Institutes of Health, Bethesda, MD, USA

Javier Gonzalez-Castillo

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Javier Gonzalez-Castillo .

Ethics declarations

Competing interests.

The author declares no competing interests.

Rights and permissions

Reprints and permissions

About this article

Cite this article.

Gonzalez-Castillo, J. Traveling and standing waves in the brain. Nat Neurosci 25 , 980–981 (2022). https://doi.org/10.1038/s41593-022-01119-0

Download citation

Published : 28 July 2022

Issue Date : August 2022

DOI : https://doi.org/10.1038/s41593-022-01119-0

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

This article is cited by

  • Alexey V. Scherbakov
  • Manfred Bayer

Nature Communications (2023)

Quick links

  • Explore articles by subject
  • Guide to authors
  • Editorial policies

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

brain travelling waves

June 28, 2018

“Traveling” Brain Waves May Be Critical for Cognition

Physical motion of neural signals may play a more important role in brain function than previously thought

By Simon Makin

brain travelling waves

Getty Images

The electrical oscillations we call brain waves have intrigued scientists and the public for more than a century. But their function—and even whether they have one, rather than just reflecting brain activity like an engine’s hum—is still debated. Many neuroscientists have assumed that if brain waves do anything, it is by oscillating in synchrony in different locations. Yet a growing body of research suggests many brain waves are actually “traveling waves” that physically move through the brain like waves on the sea.

Now a new study from a team at Columbia University led by neuroscientist Joshua Jacobs suggests traveling waves are widespread in the human cortex—the seat of higher cognitive functions—and that they become more organized depending on how well the brain is performing a task. This shows the waves are relevant to behavior, bolstering previous research suggesting they are an important but overlooked brain mechanism that contributes to memory, perception, attention and even consciousness.

Brain waves were first discovered using electroencephalogram (EEG) techniques, which involve placing electrodes on the scalp. Researchers have noted activity over a range of different frequencies, from delta (0.5 to 4 hertz) through to gamma (25 to 140 Hz) waves. The slowest occur during deep sleep, with increasing frequency associated with increasing levels of consciousness and concentration. Interpreting EEG data is difficult due to their poor ability to pinpoint the location of activity, and the fact that passage through the head blurs the signals. The new study, published in June in Neuron , used a more recent technique called electrocorticography (ECoG). This involves placing electrode arrays directly on the brain’s surface, minimizing distortions and vastly improving spatial resolution.

Scientists have proposed numerous possible roles for brain waves. A leading hypothesis holds that synchronous oscillations serve to “bind” information in different locations together as pertaining to the same “thing,” such as different features of a visual object (shape, color, movement, etcetera). A related idea is they facilitate the transfer of information among regions. But such hypotheses require brain waves to be synchronous, producing “standing” waves (analogous to two people swinging a jump rope up and down) rather than traveling waves (as in a crowd doing “the wave” at a sports event). This is important because traveling waves have different properties that could, for example, represent information about the past states of other brain locations. The fact they physically propagate through the brain like sound through air makes them a potential mechanism for moving information from one place to another.

These ideas have been around for decades , but the majority of neuroscientists have paid little attention. One likely reason is that until recently most previous reports of traveling waves—although there are exceptions—have merely described the waves without establishing their significance. “If you ask the average systems neuroscientist, they’ll say it’s an epiphenomenon [like an engine’s hum],” says computational neuroscientist Terry Sejnowski of the Salk Institute for Biological Studies who was not involved in the new study. “And since it has never been directly connected to any behavior or function, it’s not something that’s important.”

The tools researchers use may also have played a part. Today’s mainstream neuroscience has its roots in studying the behavior of neurons one at a time using needlelike microelectrodes. Pioneering researchers in this area noticed the timing of when a neuron fired varied from one trial of an experiment to another. They concluded this timing must not be important and began combining responses from multiple trials to produce an average “firing rate.” This became the standard way to quantify neural activity, but the variability may result from where neurons are in oscillation cycles, so the practice ignores the timing information needed to reveal traveling waves. “The conceptual framework grew out of what a single neuron is doing by itself,” Sejnowski says, but “the brain works through populations of neurons interacting with each other.” Because traveling waves comprise the activity of many neurons spread across the brain, they are invisible to single-neuron techniques. But over the last decade new technologies have appeared that allow many neurons to be monitored simultaneously. “This has given us a very different picture,” Sejnowski says. “For the first time we have the tools and techniques to see what’s really going on—but it’s going to take a generation before it’s accepted by the established neuroscience community.”

Optical methods, like voltage-sensitive dyes, allow researchers to visualize electrical changes in thousands of neurons simultaneously but cannot be used in humans because of the risks they pose. ECoG, however, is commonly used in epilepsy patients to investigate seizures. So the researchers behind the new study recruited 77 epilepsy patients with implanted ECoG arrays and went looking for traveling waves. They first looked for clusters of electrodes displaying oscillations at the same frequency. Nearly two thirds of all electrodes were part of such clusters, which were present in 96 percent of patients (at frequencies from 2-15 Hz, spanning the theta band at 4-8 Hz and alpha band at 8-12 Hz). The researchers next assessed which clusters represented bona fide traveling waves by analyzing the timing of the oscillations. If consecutive oscillations are part of a traveling wave, each will be slightly delayed or advanced, depending on direction of travel. (Think of how people in a crowd wave follow one another with a slight delay.) Two thirds of the clusters detected were traveling waves moving from the rear to the front of the cortex. These involved nearly half of all electrodes and occurred in all lobes and both hemispheres of patients’ brains.

The team next gave participants a working-memory task and found traveling waves in their frontal and temporal lobes became more organized half a second after people were prompted to recall information. The waves changed from moving in various directions to mostly moving in concert. Importantly, the extent to which they did this varied with how quickly participants responded. “More consistent waves correspond to better task performance,” Jacobs says. “This suggests a new way to measure brain activity to understand cognition, which can perhaps give rise to new, improved brain–computer interfaces.” (BCIs are devices that connect a human brain to a machine that performs some task, like moving a prosthetic limb.)

These findings should help dispel some researchers’ lingering doubts about the importance of such waves. “The article is a strong contribution to the study of cortical traveling waves, adding to previous work on their role in human cognition,” says psychologist David Alexander of the University of Leuven in Belgium who did not take part in the work. “This really will put to rest any worries that the waves are an artifact of blurring of signal passing through the skull.” He also says the authors make unjustified claims about the novelty of the findings and fail to acknowledge some previous research, however. “Previous work on traveling waves has shown they are evoked during working memory tasks,” he says, pointing to a 2002 EEG study that found the timing of a reversal in direction of theta waves correlated with memory performance. Interestingly, an EEG study Alexander himself published in 2009 found fewer waves moving from the front to the back of the head during a working-memory task in people who had experienced their first episode of schizophrenia, compared with healthy individuals, suggesting differences in traveling wave behavior can be related to psychiatric symptoms. He also claims the methods the team used to assess traveling waves are similar to those he used in a 2016 study . “Alexander’s work is really interesting, but it’s not clear his findings involve the same signals as our paper,” Jacobs notes. “He reported patterns that literally involve the entire brain whereas our findings were limited to particular regions.” Jacobs also points to differences in recording techniques and the nature of recorded signals.

Confirming the importance of traveling waves creates new horizons in neuroscience. “Finding that such a wide range of oscillations are traveling waves shows that they involve coordinating activity across different brain regions,” Jacobs says. “This opens key new areas of research, such as understanding what exactly this coordination consists of.” He thinks the waves propagate information, at least in the context of the current study.

Another idea holds that waves, by repeatedly moving across patches of cortex, modulate the sensitivity of neurons so as to sweep a “searchlight” of attention across, say, the brain’s visual processing area. “The concept of a traveling wave is closely tied up with the issue of how you maintain the cortex in the sweet spot where it’s maximally sensitive to other inputs and able to function optimally,” Sejnowski says. Interest in traveling waves will undoubtedly continue to increase. “What you’re seeing right now is a transformation from one conceptual framework to a completely new framework,” he adds. “It’s a paradigm shift.”

SA Mind Vol 29 Issue 5

Loading metrics

Open Access

Peer-reviewed

Research Article

Traveling waves in the prefrontal cortex during working memory

Roles Formal analysis, Writing – original draft, Writing – review & editing

Affiliation The Picower Institute for Learning and Memory and Department of Brain and Cognitive Sciences, Massachusetts Institute of Technology, Cambridge, Massachusetts, United States of America

Roles Data curation, Investigation, Resources, Writing – review & editing

ORCID logo

Roles Investigation

Roles Conceptualization, Funding acquisition, Project administration, Supervision, Writing – original draft, Writing – review & editing

* E-mail: [email protected]

  • Sayak Bhattacharya, 
  • Scott L. Brincat, 
  • Mikael Lundqvist, 
  • Earl K. Miller

PLOS

  • Published: January 28, 2022
  • https://doi.org/10.1371/journal.pcbi.1009827
  • See the preprint
  • Reader Comments

Fig 1

Neural oscillations are evident across cortex but their spatial structure is not well- explored. Are oscillations stationary or do they form “traveling waves”, i.e., spatially organized patterns whose peaks and troughs move sequentially across cortex? Here, we show that oscillations in the prefrontal cortex (PFC) organized as traveling waves in the theta (4-8Hz), alpha (8-12Hz) and beta (12-30Hz) bands. Some traveling waves were planar but most rotated. The waves were modulated during performance of a working memory task. During baseline conditions, waves flowed bidirectionally along a specific axis of orientation. Waves in different frequency bands could travel in different directions. During task performance, there was an increase in waves in one direction over the other, especially in the beta band.

Author summary

We found that oscillations in the prefrontal cortex form “traveling waves”. Traveling waves are spatially extended patterns in which aligned peaks of activity move sequentially across the cortical surface. Some traveling waves were planar but most rotated. The prefrontal cortex is important for working memory. The traveling waves changed when monkeys performed a working memory task. There was an increase in waves in one direction over the other, especially in the beta band. Traveling waves can serve specific functions. For example, they help maintain network status and help control timing relationships between spikes. Given their functional advantages, a greater understanding of traveling waves should lead to a greater understanding of cortical function.

Citation: Bhattacharya S, Brincat SL, Lundqvist M, Miller EK (2022) Traveling waves in the prefrontal cortex during working memory. PLoS Comput Biol 18(1): e1009827. https://doi.org/10.1371/journal.pcbi.1009827

Editor: Dietmar Plenz, Porter Neuroscience Research Center, National Institute of Mental Health, UNITED STATES

Received: June 21, 2021; Accepted: January 11, 2022; Published: January 28, 2022

Copyright: © 2022 Bhattacharya et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: Code and preprocessed data are available at: https://github.com/sayak66/wM_travelingwaves_code .

Funding: This work was supported by ONR MURI N00014-16-1-2832 (E.K.M.), NIMH R37MH087027 (E.K.M.), The MIT Picower Institute Innovation Fund (E.K.M.). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Oscillatory dynamics have been linked to a wide range of cortical functions. For example, higher frequency (gamma, >40 Hz) power (and spiking) increases during sensory inputs (and their maintenance) and during motor outputs [ 1 – 5 ]. Gamma power is anti-correlated with lower frequencies (alpha/beta, 8–30 Hz), whose power is often higher during conditions requiring top-down control (e.g., when attention is directed away, or an action is inhibited) [ 6 – 8 ]. Such observations have led to a theoretical framework in which oscillatory dynamics regulate neural communication [ 9 – 11 ].

Thus far, most studies of neural oscillations have focused on what we can call “standing wave” properties (e.g., power of and coherence between oscillations at different cortical sites), ignoring any organization of where and when the peaks and troughs of activity appear. However, there is mounting evidence for such organization. Oscillations can take the form of “traveling waves”: Spatially extended patterns in which aligned peaks of activity move sequentially across the cortical surface [ 12 , 13 ]. This apparent movement of the amplitude peak is facilitated by the existence of a phase gradient along a particular direction, along which the movement occurs. Traveling waves have most often been reported in the lower-frequency bands (<30 Hz). Examples include beta-band (15–30 Hz) traveling waves in motor and visual cortices [ 14 , 15 ] and theta band (3–5 Hz) traveling waves in the hippocampus [ 16 , 17 ].

Traveling waves are of interest because they have a variety of useful properties for cognition, development, and behavior. They can create timing relationships that foster spike-timing-dependent plasticity and memory encoding [ 14 , 18 ]. They add information about recent history of activation of local networks [ 19 ]. They are thought to help “wire” the retina [ 20 ] and cortical microcircuits during development [ 21 ]. Their functional relevance in the adult brain is suggested by observations that traveling wave characteristics can be task-dependent and that they impact behavior. For example, EEG recordings have shown that alpha band waves reverse their resting state direction during sensory inputs [ 22 ]. Behavioral detection of weak visual targets improves when there are well organized low-frequency (5-40Hz) traveling waves in visual cortex vs when there is a weaker, “scattered” organization [ 23 ].

Oscillatory activity in the prefrontal cortex has been linked with cognitive functions like working memory and attention but there has been little examination of whether they form spatio-temporal structures like traveling waves. Most studies have averaged oscillations across spatially distributed electrodes. This increases the “signal” of the oscillations but prohibits analysis of any spatial organization. Thus, we examined their spatial organization from microarray recordings in the PFC of monkeys performing a working memory task. This revealed that low-frequency (beta and lower) traveling waves are common in the PFC. We characterized their speed, direction and their patterns. They often rotated and changed direction during performance of a working memory task.

Two animals performed a delayed match-to-sample task ( Fig 1A ). They maintained central fixation while a sample object (one of eight used, novel for each session) was briefly shown. After a two-second blank memory delay (with maintained fixation), two different objects (test screen, randomly chosen from the eight) were simultaneously presented at two extrafoveal locations. Then the animals were rewarded for holding fixation on the object that matched the sample. Two 8x8 multi-electrode “Utah” arrays, one in each hemisphere, were used to record local field potentials (LFPs) from the dorsal pre-frontal cortex (dlPFC). All data is from correctly performed trials. We analyzed data from 14 experimental sessions, five from one animal (Animal/Subject 1) and nine from the other (Animal/Subject 2).

thumbnail

  • PPT PowerPoint slide
  • PNG larger image
  • TIFF original image

(A) Delayed-match-to-sample (DMTS) working memory task. The subject fixated at the center for 0.5 s (leftmost panel) before an object (one of eight) were presented at the center for another 0.5 s. There was a 2 s blank memory delay after which a test screen was presented at extrafoveal locations. The subject had to saccade towards the remembered object and hold fixation there (arrow, rightmost panel). (B) Filtered LFP power trends for all electrodes (four arrays), trial-averaged and shown across time with the lines denoting the major trial epochs. Four frequency ranges were chosen–theta (4–8 Hz), alpha (8–12 Hz), beta (12–30 Hz) and gamma (40–120 Hz), with the dots above the curve denoting if the LFP power at that instant is significantly (p<0.01) different from baseline (0.5 before fixation started). (C) Theta LFP organization across the right hemisphere array of Subject 1 of a particular trial. Each tile denotes a recording site on the array (8x8 total) with the color denoting the LFP amplitude at that site. The array position with respect to anatomical brain landmarks is overlaid. Each panel denotes the organization at an instant in time. The black arrow in the second panel indicates the direction of movement of the high amplitude LFPs with time. (D) Voltage traces from the 8 adjacent electrodes (color graded by position) in a row of the array in (C–dashed line). The circles mark the peak positions of the oscillation cycles with the dashed line indicating how the peaks shift gradually in time and space. (E) Phase maps from the corresponding panels in (C) with the color on each tile denoting the phase of the LFP oscillation cycle on that recording site.

https://doi.org/10.1371/journal.pcbi.1009827.g001

LFP oscillations formed traveling waves

Task-related modulations of LFP power were similar to that seen in previous studies. Fig 1B shows average changes in LFP power from baseline (a 500ms period just before fixation) across all electrodes, all sessions, and both animals as a function of time during the trial. Alpha (8–12 Hz) and beta (12–30 Hz) power decreased during sample presentation and increased during the delay relative to baseline. By contrast, theta (4–8 Hz) and gamma (40–120 Hz) power showed opposite trends. They increased during sample presentation and test screen presentation. Gamma power also ramped up near the end of the delay.

These increases in oscillatory power were not equal or simultaneous across the electrode array. Rather, there was a spatio-temporal structure that suggested traveling waves of activity. An example of a beta-band (12–30 Hz) wave during test screen presentation of one trial is shown in Fig 1C ( S1 Movie ). The tiles represent the electrodes in the array. The array is oriented relative to its position with respect to anatomical brain landmarks in the dlPFC. The colors show the beta-band signal amplitude at each electrode at different snapshots in time. Note that, at first (t = 0), beta is higher near the top of the array. Then, over time, the higher beta amplitude shifts systematically to the bottom of the array. To illustrate this in another way, Fig 1D shows the filtered LFP traces across one row of the array (dashed line in Fig 1C ; the electrodes are plotted on the Y-axis from left to right ascending). The oscillation peaks are marked. A sequential shift of the LFP peak activation across the row of electrodes can be seen. This shifting peak of activity in space with time is characteristic of traveling waves.

As a traveling wave moves across cortex, it should produce a phase gradient across adjacent recording sites. This can be seen in Fig 1E . It replots the data from Fig 1C as a phase map. The instantaneous beta-band phase at each electrode is color-coded for each time slice. Zero phase is the peak of each oscillation. Positive values indicate LFPs that were on their way to peak amplitude. Negative values indicate LFPs that have passed the amplitude peak. Note the phase gradient across the array and how it changes over time. At first, positive phase values tend to be higher on the bottom of the array. Then, as the wave travels along the arrow, the phase values on the top become negative (indicating that they are past the peak of the wave), phase values in the bottom shift toward zero over time (indicating the wave approaching peak) followed by a shift to negative values. But, at each point in time, the phase gradient consistently points in the direction of wave propagation. This can also be seen in Fig 1D . At the timepoints when the topmost electrode is at peak phase, the rest of the electrodes form a gradient down the rising phase of the sinusoidal wave. Thus, the instantaneous phase gradient provides a signature of a traveling wave and its direction at each point in time during a wave traversal.

A traveling wave was counted when there was a strong phase gradient in a particular direction. We counted the number of traveling waves across all sessions and both arrays in both animals by computing the trial-by-trial correlation between the observed phase map at each time point and a Euclidean distance map (with increasing distance in the direction being checked, see Materials and Methods and S1A Fig ). Quadrant-based distance maps were used to calculate waves across twelve directions (all pairwise combinations of four quadrants × two directions, see Materials and Methods ). When the correlation between the distance and phase map exceeded a certain threshold, a wave in that direction was counted. This threshold was set such that detected waves were required to have correlations greater than 99% of those obtained by random shuffling of the electrode locations in the array (see Materials and Methods ). We further used simulations to validate our wave classification algorithm ( Materials and Methods ). We applied this classification across time in each trial using a shifting 20ms time-interval (short enough to capture anything less than 50Hz), averaging across trials and across all four arrays in the two animals.

Waves were apparent throughout the trial, albeit at different times for different frequency bands. Fig 2A shows the wave count (summed across all directions) for the theta, alpha, and beta frequency bands. For all bands, there were consistent increases in wave counts, relative to baseline (a 500ms window before fixation), during fixation, sample and test screen presentations ( Fig 2A ). They elicited bursts of sharp, consecutive increases and decreases in wave prevalence over time (i.e., significantly above and below baseline values corresponding to a time-locked traveling wave followed by brief pause).

thumbnail

(A) Number of wave components observed in the three frequency ranges during the trial. The numbers are trial-averaged across all four arrays. The dots underneath denote if the numbers at that instant are significantly different from baseline (0.5s before fixation). (B) Wave speeds observed across all frequency ranges. Error bars represent SEM. Star indicates statistically significant difference (p<0.01).

https://doi.org/10.1371/journal.pcbi.1009827.g002

The wave trends over time ( Fig 2A ) mirrored the power modulations shown in Fig 1B . For the theta and alpha bands, there were significantly higher-than-baseline wave counts in all epochs. Counts peaked during sample presentation (for theta) and test screen presentation (for theta and alpha). Beta waves also increased relative to baseline in all epochs. However, in contrast to alpha and theta bands, beta traveling waves tended to be higher during the memory delay and decreased during sample and test screen presentation. Wave speed ranged from 20–60 cm/sec and increased with increases in frequency band ( Fig 2B ) as expected [ 15 ].

Rotating waves

We found that the waves often had a rotational component. To determine this, we examined the circular-circular correlation coefficient (ρ c ) between observed phase maps and rotation maps across different points of the array (see Materials and Methods ). This represented the circular correlation between a rotation map centered around a point on the array and the array phase map for that slice of time. The correlation provided an estimate of the wave direction vector. For example, in Fig 3A , the red arrows showed positive coefficients for the central correlation coefficient (see Materials and Methods , S1B Fig ).

thumbnail

(A) Schematic illustrating the function of the circular-circular correlation coefficient (ρ c ) around a particular point (yellow circle). Two different positions of the point are shown. A positive ρ c is calculated for the waves toward the red part of the array (red vectors), while blue gives a negative ρ c . The grey zone denotes wave directions that gave a correlation coefficient value within the “chance zone”. The zones differ with choice of point (right vs left panel) (B) Two broad wave types–planar (left) and rotating (right) waves were classified. Finer distinction possibilities are also shown–different directions (planar) or different spatial wavelengths (rotating). A combination of three coefficient values were used to distinguish between these wave types. The grey circles denote the points on the array around which the coefficient was calculated. (C) Plots showing the number of time instants in which rotating/planar waves were observed across all arrays and trials in the three frequency ranges. Star indicates statistically significant difference (p<0.01). (D) Plots showing the different wavelengths observed for the rotating waves classified in (C). (E) Wave directions (planar and rotating combined) observed across all trials in Subject 2 (left hemisphere) overlaid on the array position with respect to brain landmarks. The three colors denote the three frequency ranges. (F) Wave directions similar to (E), but for all four arrays combined, with each array’s most preferred direction aligned along the 0-degree axis.

https://doi.org/10.1371/journal.pcbi.1009827.g003

A range of ρ c captured a spectrum of wave types, from planar to rotating waves, of different directions and spatial wavelengths ( Fig 3B , see Materials and Methods ). A high ρ c value indicated a wave moving in a particular direction–but did not distinguish between the type of wave (planar vs rotating, S1C Fig ). To distinguish between planar and rotating waves, we used the circular-circular correlation coefficient across three points of the array (shown by dark circles in Fig 3B ). Each of these coefficients had a distinct sensitivity profile to rotational and planar waves. Thus, the combination of the three ρ c values provided a fingerprint-like signature that could be used to accurately classify wave types (planar vs rotating) and identify their wavelength and direction ( Materials and Methods ). Simulations were used to quantify ρ c around the three points for different types of planar and rotating waves (see Materials and Methods , S1D and S1E Fig ). The set of coefficients computed for the LFP data was then compared with the simulation sets to classify waves based on which set was closest to the LFP coefficients in terms of Euclidean distance ( Materials and Methods ). Fig 3C shows the number of planar vs rotating waves observed across all four arrays. Overall, rotational waves were more common than planar waves for all frequency bands in all task epochs. An example theta rotating wave (Subject 1, right dlPFC, memory delay) is shown in S2 Movie . This method was not constrained by the exact choice of points on the array ( Materials and Methods ). Analyzing ρ c values around other points on the array ( S2A Fig ) showed similar results.

Our arrays were relatively small (3x3mm), meaning that it was possible, if not likely, that our arrays were capturing pieces of a larger rotating wave structure. A rotating wave will show different spatial wavelengths (i.e. the spatial distance from one wave peak to the next) depending on the position of the center of the rotating wave relative to the array. As one moves away from the core of the rotating wave, the spatial wavelength increases. We quantified the incidences of waves of different spatial wavelengths ( Fig 3D ). For all frequency bands, we observed a greater incidence of waves with longer wavelengths. This suggests that, indeed, we were capturing pieces of a larger rotating structure. We will return to this point later.

Waves travelled in preferred directions

The waves did not travel in random directions. To check if there were preferred directions of travel along each array, we leveraged a property of the circular-circular coefficient. The coefficient could discriminate between waves in opposing directions (red and blue regions respectively, Fig 3A ) along a particular axis. Waves directed towards the positive half (red arrows, Fig 3A ) had ρ c >0 and vice versa (blue arrows). The orientation of the axes splitting the positive and negative regions depended on the net direction of the rotation map chosen and hence differed with the chosen point on the array ( Fig 3A , left vs right, the chosen point shown in yellow). We chose points such that we could split the array into polar segments of around 10–15 degrees each. We confirmed these properties with simulated data (see Materials and Methods ). To test for statistical significance of the correlation obtained, we calculated a ρ c threshold beyond which the phase organization exceeded chance (p<0.01, ρ c >0.3, random permutation test). This approach allowed us to measure wave directions but did not discriminate between planar and rotating waves.

Fig 3E shows the placement of one of the arrays relative to the principal and arcuate sulci, with polar histograms of observed wave directions overlaid. Some wave directions were more common than others. This was consistent across the frequency bands. We reoriented the data from all four arrays such that their most preferred direction (for each frequency band) was along the horizontal axis ( Fig 3F ). The wave counts in each direction were then averaged across trials. Clear directional preferences were seen across all arrays and all frequency bands (theta to beta). Note that the directional preference remained consistent across trials epochs including the baseline (dashed lines). In other words, there were preferred “default” directions. Task performance increased or decreased the probability of waves traveling in those directions. A degree of bimodality was also observed for all frequencies. In addition to the preferred direction at 0° (by definition), there was a secondary preference for directions around 180°. Thus, there was typically a preferred axis of wave propagation (dashed arrow, Fig 3F ) with an additional preference for one direction over the other along that axis.

Task-related changes in wave direction

Though task demands did not change the preferred axis of wave motion, we found that it could alter the balance between opposite directions along that axis. To determine this, we again used the circular-circular correlation analysis. For this purpose, we examined correlation values (ρ c ) with the rotation map around the point that showed the maximum number of waves for each array, adjusting such that the task-enhanced direction had positive ρ c . This approach included both planar and rotating waves. Oppositely directed waves showed opposite signs.

Fig 4A shows the distribution of ρ c values, averaged across all trials and all four arrays in both animals. Results are shown separately for three frequency bands, and for three task epochs (green lines), with each compared to the pre-trial baseline (red lines). The ρ c at each time instant was classified as a wave if it exceeded the value expected by chance (indicated by shaded areas of Fig 4A , see Materials and Methods ). Alpha and theta waves showed unimodal ρ c histograms centered around zero during baseline and delay epochs. Presentation of the sample or test-array shifted the histogram peak out of the shaded area (indicating increase in wave incidence) but did so in one direction over the opposite. This could be seen in theta and alpha frequency bands ( Fig 4A ).

thumbnail

(A) Histograms quantifying the correlation values seen on average in each trial during the sample, delay, and test-onset intervals (all 0.5 s in length), combined across arrays. A correlation value to the right of the shaded region (positive) denotes waves in a particular direction, while the left means the opposite direction. The shaded region denotes the “chance zone” where no conclusion regarding wave direction can be made. For each frequency range (each row), the red histogram corresponds to the correlations observed in baseline conditions (0.5 s pre-fixation), while the green histogram corresponds to the correlations observed in that epoch (0.5 s). The blue dots denote if the two are significantly different from each other (p<0.01). (B) Quantification of the difference between the green and red curves in (A) for all 0.5 s intervals during each trial. The red line shows difference from baseline for the positive wave direction, while the blue line for the negative wave direction.

https://doi.org/10.1371/journal.pcbi.1009827.g004

Beta waves showed prominent bidirectionality throughout the trial ( Fig 4A , bottom row). During both the baseline (red line) and task performance (green line), the distribution of ρ c values for waves in the beta band were bimodal with “bumps” on the ends of the distributions (outside the “chance zone”). This indicates waves in opposite directions. Relative to baseline, during task performance the “bump” on one end of the distribution rose while the other lowered, indicating an increase in waves in one direction and a decrease in the opposite. Baseline ρ c values (red lines) for the beta band skewed toward the negative direction indicating a baseline default bias for waves to travel in that direction. During the sample and delay epochs (green lines), waves skewed more toward positive ρ c values). After test screen presentation and the monkey’s behavioral response, this reversed back to a skew toward negative ρ c values, the baseline direction opposite of that seen in the task.

This is illustrated in more detail in Fig 4B . It shows changes in wave direction bias (relative to baseline) over time. For theta and alpha waves, negative ρ c values were significantly increased from baseline during the sample (blue line, Fig 4B ) and especially after the test screen appeared. Positive ρ c values also increased intermittently for these two frequency bands. The beta band showed the most consistent changes in wave direction preference with task performance ( Fig 4B , bottom row). During sample presentation and the memory delay, there was an increase of positive ρ c values and a decrease in negative ρ c values indicating a consistent shift toward waves flowing in one direction over the other. After test screen presentation and the animal’s behavioral response (i.e., post-test), this shift ended. There was a decrease in positive, and an increase in negative ρ c values resulting in the mix of the two directions seen during baseline.

For this representation, we adjusted the coefficient points such that the enhanced wave direction had positive ρ c for all frequencies . However, it is important to note that although the waves in different frequency bands had similar preferred direction axes ( Fig 3F ), this did not mean that the waves in different bands traveled in the same directions at the same time, as part of a multiband wave. S3 Fig shows the correlation coefficient calculated around the central point (4,4) for the left dlPFC array of Animal 2. As can be seen in the histograms ( S3A Fig ), theta waves preferred the negative direction during baseline, sample presentation and delay (higher histogram bump towards negative ρ c values). By contrast, beta waves preferred the opposite direction, i.e., the higher histogram bump was towards positive ρ c values. This is also evident in the quantification of the directional increase during task performance from baseline ( S3B Fig ). While the negative direction was enhanced during the task for theta waves, positive values were enhanced for beta waves.

Inferring the larger wave structure from observed dynamics

For the analyses above, we aligned the waves along the preferred axis/direction of the waves for each array. In reality, the waves flowed in different anatomical directions across each individual array. This was likely due to different placement of each array relative to a larger wave structure. We used features of rotating waves to infer that larger structure.

Rotating waves create a heterogeneous vector field with the following features (illustrated in S1D and S1E Fig ): 1. The local movement of waves from the same rotating structure will flow in different directions in different areas around the center of rotation. 2. Rotating wave organization decays as one moves away from the center causing the wave to lose structure. 3. The spatial wavelength of the wave (the spatial distance from peak to peak) grows as one moves away from the center of rotation. Thus, shorter wavelengths on the array indicated that the rotating wave center was closer to the center of the array and vice versa. We also used our correlations coefficients to quantify the exact direction and type of wave pattern ( S2B and S2C Fig ). We estimated the locations of the rotating waves using these two metrics. The three frequency bands had similar wave directions and curvatures. However, the beta band tended to show the highest traveling wave counts (e.g., in Fig 4A ). Thus, to make better inferences, we focused on the beta band.

The results were consistent with our arrays capturing parts of two oppositely rotating waves. The rotating waves seemed to be located similarly in each hemisphere for both subjects relative to anatomical brain landmarks (the arcuate and the principal sulci). Fig 5A shows the spatial wavelengths recorded for Animal 1, along with the directional histograms for beta-band waves. The array in the right hemisphere recorded more short-wavelength waves than the array in the left hemisphere. The result is a wider ρ c histogram spread and stronger directional differences than the left hemisphere array. This would suggest that the rotating wave centers were closer to the array in the right hemisphere, i.e. the recording array captured more of the rotating waves in this hemisphere. Based on the wave features described above (shown as bar graphs), Fig 5B shows hypothesized locations of the larger rotating wave for each rotational direction relative to anatomical landmarks in each hemisphere (see Materials and Methods ).

thumbnail

(A) Quantification of the differences observed in beta-band waves between the left and right hemispheres of Subject 1 with regard to spatial wavelengths observed (left), correlation coefficients (center), and wave directional differences (right). (B) Quantification of the levels of different wave types (bar graphs, with the wave type indicated through arrows below the corresponding bar graph) on each array–overlaid with respect to the array position in the brain–for both left and right hemispheres of Subject 1. The red and blue rotating directions indicate the inferred locations of the opposite rotating waves based on the pattern of different wave types observed. (C,D) Same as (A) and (B) but for Subject 2.

https://doi.org/10.1371/journal.pcbi.1009827.g005

Fig 5C shows similar analysis for Animal 2. In this case, the array in the left hemisphere showed significantly stronger effects than the right, indicating the position of the left hemisphere array was closer to the center of rotation. For example, the array in the left hemisphere of Animal 2 recorded the highest number of short spatial wavelengths. Thus, it can be hypothesized that the rotating wave covered the most array area compared to the other three arrays.

Additionally, we found a correlation between spike rate and rotating wave organization. Spike rates were higher closer to the center of the rotating wave ( S4 Fig ). Shorter wavelengths indicate that the center of the rotating wave is closer to the array ( S4A Fig ). We found that shorter wavelengths were associated with higher spike rates ( S4B Fig ). This indicates that rotating waves modulate spike rates.

We found that 4–30 Hz (theta, alpha, and beta band) oscillations in the lateral prefrontal cortex organize into traveling waves. Rotating waves outnumbered planar waves. Average spike rates were higher near the rotating wave centers. Before the trial began, traveling waves had a preferred orientation but tended to be bidirectional, flowing in opposite directions randomly. After behavioral trial initiation, there was an increase in waves in one direction. This was especially evident during presentation of a to-be-remembered sample and subsequent presentation of a test screen of two stimuli. Notably, in the beta band, there was a persistent increase in the flow of waves in one direction while waves in the opposite direction decreased during the delay. Our animals were well trained on the task showing correct responses in 87% of the trials. This did not give us enough statistical power to compare traveling wave characteristics in correct vs error trials. Future studies will be needed to address this.

The neural infrastructure of traveling waves

The existence of traveling waves provides insight into the underlying circuitry. They can be explained by short-range lateral interactions between weakly-coupled oscillators [ 12 ]. This creates phase gradients, i.e. time lags, between successive regions resulting in a sequential movement of a peak of activity.

We found that the waves flowed back and forth in opposite directions. This could be explained by activation of different subsets of neurons with different wave direction preferences. Increasing activation of one network could inhibit the other thus biasing flow in one direction [ 24 ], as we observed during task performance especially in the beta band. Removal of activation from this network and the corresponding release of inhibition from the other network could explain our observation of a decrease in preferred, and an increase in the non-preferred, wave direction after the trial ended.

The mechanism of generating rotating waves is a subject of extensive theoretical research [ 25 ]. A rotating wave is different from a planar or radial wave. Rotating waves create a heterogeneous vector field, i.e., a different phase-gradient depending on where one is recording [ 26 ]. It has altered curvature along the arm of the wave. That is, different points along the wave have different curvature, hence different speeds [ 27 ], rendering it an extra degree of freedom when compared to planar waves. Rotating waves are formed when spatial heterogeneities lead to the formation of phase-singularities. This heterogeneity is mostly caused by a core circuit with altered excitation/inhibition levels around which the wave starts to rotate [ 28 ]. How this altered core forms varies from field to field [ 29 , 30 ], and in the context of neural mechanisms, remains an open question.

While most studies have reported planar travel waves; a few, like ours, also found rotating waves [ 26 , 31 ]. Rotating waves span wide expanses of human cortex during sleep spindles, appear in visual cortex of turtles following sensory stimulation, or spontaneously in rodent cortical slices (visual cortex). But as we demonstrated in our analysis, the location of the recording arrays relative to the wave affects observations. Rotating waves also tend to drift around in space, further complicating the consistent capturing of the rotating center on the recording array [ 28 ]. This is consistent with analyses suggesting that some planar waves might be incomplete observations of parts of rotating waves [ 32 ]. More accurate assessment of the relative proportions of rotating vs planar waves can be made with larger recording arrays.

Functional role of rotating/traveling waves in working memory

One advantage of traveling waves over synchronous oscillations (standing waves) is in maintaining current network status/information. Standing waves result in time periods when all neurons in a network or subnetwork are turned “off”. By contrast, traveling waves ensure that a subset of a given network is always “on” [ 32 ]. Maintaining “on” states is particularly pertinent for holding items in working memory. Indeed, a recent modeling study shows that segregating memories and maintaining them in neural modules is more robust and stable when the oscillations are phase-shifted among neural ensembles, i.e. manifest as traveling waves [ 33 ]. Although traveling waves were dominant ( S5A Fig ), “standing-wave”-type oscillations were also observed, albeit in much lower numbers. It is important to note that a perfect standing wave (no phase variance) is unlikely. Oscillations could be considered “standing” when there was no clear phase gradient and the phase variance was lower than for a traveling wave ( S5 Fig ).

Traveling waves may also play a role in a fundamental cortical function: predictive coding. The brain continually generates predictions of immediately forthcoming inputs, to prevent processing of predicted (thus uninformative) inputs, preventing sensory overload [ 34 ]. Unpredicted/new inputs are allowed to pass as “prediction errors”. A model of the cortex built by VanRullen and colleagues [ 22 ] shows how alpha-band traveling waves carry “priors” from higher to lower cortex in the absence of sensory inputs. Sensory inputs evoke traveling waves that flow in the opposite direction.

Stimulus-evoked traveling waves can add information about time and recent activation history to the network. Muller et al. 2018 [ 19 ] suggest that traveling waves allow temporal reversibility: One can decode the history of a pattern from the observed spatiotemporal structure. A given pattern of network activations sets up a unique pattern of traveling waves. This temporal sequence can be decoded to track the elapsed time and recent activation patterns. Elapsed time was not behaviorally relevant in our working memory task; animals were cued when to respond. But relevant or not, the brain keeps track of time. A common example is PFC spiking ramping up near the end of a fixed delay interval in expectation of the forthcoming decision or behavioral response [ 9 , 35 – 37 ]. Recent work has shown PFC neurons track time via neurons that spike at particular times during a memory delay [ 38 ]. Traveling waves may provide a mechanism of such tracking, through phase-based temporal encoding–i.e. assigning a unique phase map to each instant of time. Tracking time potentially allows neurons to be prepped for future events rendering the cortex with predictive abilities.

We did not observe any changes in traveling wave with different sample items. This is consistent with theories that traveling waves have “meta” network functions independent of the items or other content being processed as mentioned above.

We found higher firing rates closer to the center of the rotating wave. This modulation may provide computational advantages over its planar counterparts. Rotating waves have been proposed to play a role in organizing and inducing neural plasticity. The repetitive dynamic of the wave caused by the rotating arm coming around in precise time-intervals can create a structure of timing differences in neural activation that induce spiking timing-dependent plasticity (STDP) in specific subsets of a network. A human ECOG study has shown rotating waves during sleep-related memory consolidation with timing differences in STDP range [ 26 ]. They showed that the coordinated organization of traveling waves is vital in maintaining plasticity. Short-lived (< 1 sec) increases in synaptic weights induced by spiking may aid in WM maintenance [ 39 ]. Thus, the waves we observed in the PFC may play a role in not only inducing this plasticity but also periodically refresh them (with spiking) so that memories can be held beyond the time constant of the short-term plasticity [ 40 ].

Neural oscillations in the low frequency ranges have been implicated in a variety of functions, including working memory, attention, and predictive coding [ 41 – 46 ]. Here, we show that they manifest in the prefrontal cortex as traveling waves that change with task performance. Traveling waves have multiple characteristics–such as direction, phase organization, speed–all of which can serve specific functions, making these waves a potentially powerful computational tool. Given their functional advantages, a greater understanding of traveling waves should lead to a greater understanding of cortical function.

Materials and methods

Ethics statement.

All procedures followed the guidelines of the Massachusetts Institute of Technology Committee on Animal Care and the National Institutes of Health (protocol 0619-035-22, approved by MIT’s Committee on Animal Care on 6/23/21).

Subjects, task and LFP recordings

The nonhuman primate subjects in our experiments were two adult males (ages 17 and 8, for Subject 1 and 2 respectively) rhesus macaques ( Macaca mulatta ).

Subjects performed a delayed match-to-sample working memory (WM) task. They began task trials by holding gaze for 500 ms on a fixation point randomly displayed at the center of a computer screen. A sample object (one of eight randomly chosen) was then shown for 500 ms at the center of the screen. After a 2 s delay, a test object was displayed. The monkeys were required to saccade to it if it matched the remembered sample. Response to the match was rewarded with juice. All stimuli were displayed on an LCD monitor. An infrared-based eye-tracking system (Eyelink 1000 Plus, SR-Research, Ontario, CA) continuously monitored eye position at 1 kHz.

The subjects were chronically implanted in the lateral prefrontal cortex (PFC) with two 8x8 iridium-oxide “Utah” microelectrode arrays (1.0 mm length, 400 μm spacing; Blackrock Microsystems, Salt Lake City, UT). Signals were recorded on a Blackrock Cerebus. Arrays were implanted bilaterally, one array in each dorsolateral PFC. Electrodes in each hemisphere were grounded and referenced to a separate subdural reference wire.

LFPs were band-passed from 0.5-300Hz and sampled at 1kHz. There were two filter stages, first a real-time analog filter in the amplifier, then a symmetric digital filter offline in software:

  • High-pass: 0.3 Hz 1 st order Butterworth
  • Low-pass: 7.5 kHz 3 rd order Butterworth
  • Low-pass: 300 Hz 3 rd order Butterworth

All correctly performed trials were included in analyses. All preprocessing and analysis were performed in Python or MATLAB (The Mathworks, Inc, Natick, MA). For the power analysis, the resulting signals were convolved with a set of complex Morlet wavelets .

LFP spatial phase maps

The raw LFP traces were filtered in the desired frequency range, using a 4 th order Butterworth filter, forward-reverse in time to prevent phase distortion (see MATLAB function filtfilt ) and interpolated for missing electrodes. Out of 28 sessions, five had missing electrodes (one electrode missing in 3 sessions, three electrodes missing in 2 sessions). Linear interpolation was done to account for the missing data. Overall as these numbers were a minority, our statistics were not affected. A Hilbert transform was used to obtain the analytical signal for each electrode. The phase of each electrode for the 8x8 array is called the “phase map” for that time instant. These phase maps (unsmoothed) were checked for gradients to identify traveling waves.

Traveling wave identification through linear distance maps

Traveling waves were identified at an instant in time when spatial correlation between the phase map at that instant and a Euclidean distance map template ( S1A Fig ) in a particular direction exceeds a certain threshold. Pearson correlation coefficients were computed over the full 2-dimensional array of values. The threshold was decided through a shuffling permutation procedure [ 26 ].

The distance map was made for each quadrant of the array, in the particular direction being evaluated ( S1A Fig shows the phase map for a diagonal wave for one quadrant). The distance maps were made such that they always increased towards the edges. So, a wave entering the array from an edge would encounter a direction map in the opposite direction–and hence show a negative correlation, while a wave leaving the array would produce a positive correlation, as it increased towards the edges. A wave thus was counted when one quadrant showed a positive correlation with the phase map, while another showed a negative correlation. Then a wave was counted to have passed from the negative quadrant to the positive quadrant. This method thus identified a wave component in a particular direction–from one quadrant to the other. Four quadrants yielded a total of 12 directions. This quadrant-based method allowed identification of wave fragments even when waves did not have uniform spatio-temporal structure throughout the whole array. So, when one wave passed, multiple direction components could be recorded. This method was verified using simulated waves.

Only positive phase values were considered in this method, so as to identify consecutive bursts of traveling wave instances. That is, when two consecutive waves passed, this method showed two peaks (two positive cycles), instead of one continuous high value for both waves. As a consequence, however, this method was restricted to identifying waves that had a long spatial wavelength when compared to the size of the array, to allow for instances where only positive phase values are on the array at one time. This approach was only used for planar wave detection. This approach is consistent with earlier studies that suggest that the spatial wavelength of traveling waves is long compared to the LFP recording array size [ 23 ].

Wave speed calculation

Wave speed at a time instant was calculated from the phases ( p ) by dividing the temporal frequency (∂ p /∂ t ) at that time with the spatial frequency (∂ p /∂ x ) [ 16 ]. The gradients obtained were averaged across electrodes to get the net wave speed for that time instant. Essentially, this calculated, how quickly the oscillation cycle progressed in space versus in time. For a fast wave, points away from each other on the array will reach the same phase within a shorter time period, i.e. a shallow spatial gradient (same ∂ p for a larger ∂ x in a certain ∂ t ), when compared to a slower wave. Wave speed was calculated only for waves of long wavelength (majority of wave instants) as spatial frequencies for short wavelength waves could vary along the array.

Shuffling procedure

To ensure that the probability of detecting traveling waves exceeded that expected by chance–we performed a random shuffling procedure to establish a threshold for the correlation coefficient–beyond which a traveling wave was counted. This was done by shuffling the phase values on the array randomly (with 25 different types of random permutations) and calculating the correlation coefficient. The 99 th percentile of the resulting distribution of coefficient values determined a threshold (0.3) above which the correlation exceeded chance.

It is important to mention here that characterizing traveling waves through quadrant-based methods or patch-based methods [ 47 ] for small arrays may lead to erroneous conclusions especially for waves with variable wavelengths or with variable local dynamics. For that purpose, we used our quadrant-based approach only for long wavelength waves which had a consistent structure on the array. To classify rotating wave patterns, we shifted to circular statistics using the whole 8x8 array.

Circular-circular correlation coefficient–planar waves, rotating waves, and net wave directionality

We used circular statistics to identify wave patterns. Circular-circular correlation coefficients have been used to classify waves in earlier studies [ 26 ]. In this study, we demonstrate that the use of circular-circular correlation coefficients can be extended to distinguish between a spectrum of wave types.

brain travelling waves

Net direction and bisecting axes.

A rotation map around a particular point has a particular net direction that can be revealed by plotting the phase maps, adjusting for circular values (i.e. subtracting the circular mean and sine transforming). The overall gradient of a phase map can best be appreciated by looking at the circular-adjusted maps. S1B Fig shows the net directions (solid lines) for rotation maps around three points. This was obtained by summing the gradients in phase across the array as complex vectors–the resultant vector being the net direction. It can be appreciated from this figure that while actual phase values may range anywhere between -pi to pi, the circular adjustment reorients all values from -1 to 1. A positive coefficient value could be obtained by multiple types of waves–rotating waves centered near the chosen point in the direction of the rotation map, or planar waves moving towards the map’s net direction . Of course, opposite directions for both would yield negative values. The orthogonal direction would result in a coefficient close to zero–which creates the bisecting axis . This approach thus allowed us to classify wave directions into positive and negative halves separated by this bisecting axis ( Fig 3A )–taking into account both rotating and planar waves. The axis was dependent on the rotation map, which varied with the chosen point.

Similar to the earlier wave classification–a correlation threshold was chosen below which no conclusions regarding wave directions could be made. This threshold was identified through shuffling the electrodes and ensuring phase gradients did not appear by chance. The 99 th percentile of the resulting distribution of coefficient values determined a threshold (0.3) above which the correlation exceeded chance. This method allowed us to categorize waves into two opposite directions ( Fig 4 ) and analyze how the directions changed during task performance. For Fig 4 , rotation maps were chosen for each array in a way that maximized the number of waves observed for that array, to provide the greatest power to discriminate between opposing directions of wave motion (Subject 1: right array–(4,4), left array–(8,4); Subject 2: right array–(4,1), left array—(4,4)).

Planar vs rotating waves.

To distinguish between planar and rotating waves, however, just one rotation map would not suffice, as the values for a rotating wave and a planar wave with the same net direction could be similar. The coefficient value using a rotation map around a particular point essentially quantifies the gradient vectors seen in that phase map. Considering, for example, a diagonal planar wave from the bottom left to the top right ( S1C Fig , left). The central point (4,4) has a net direction along that diagonal and hence records a high value around 0.9. The point (4,1) has a net direction along the horizontal and records an intermediate value around 0.64. Now, let’s consider a wave that rotates from bottom left to top-right ( S1C Fig , right). The circular-reoriented phase maps for the two waves are shown below. It is evident that the increase toward the top right is still maintained causing the coefficient around (4,4) to remain the same. However, the increase toward the horizontal (towards the right) is less when compared to the diagonal wave case. This reduces the coefficient around (4,1) to 0.52. Thus, while one coefficient could not distinguish the types of waves, a combination of two coefficients in this case was able to do so.

Wave classification, wavelength and wave structure.

brain travelling waves

Using the above simulation equation, we created a dataset comprised of 32 types of planar waves in all directions, each separated by around 12 degrees, and 40 types of rotating waves of different curvatures and wavelengths ( k ranging from 0.1 to 0.9) ( S1E Fig ). As evidence for the property described in the section above, if one monitors the (4,4) red curve in this figure, it is clear that values change signs between waves directed towards the bisecting axis (top-left to bottom-right diagonal in this case).

A long spatial wavelength (inverse of wavenumber k ) was chosen for planar waves simulations ( S1D Fig , top), as their wavelength remains constant throughout the wave band. But, rotating waves show different wavelengths at different portions of the wave ( S1D Fig , bottom). As one moves away from the rotating center (white circle), the wavelength increases. Hence, multiple wavelengths were considered while creating the datasets for rotating waves. It is also evident from S1E Fig that, in the case of rotating waves, as spatial wavelength increases (lower k )–all coefficients start to decrease–proving that the wave starts to lose spatio-temporal structure away from the rotating center. An average of 20 simulations (with distinct realization of Gaussian noise) were used to calculate the coefficient for each type of wave. We ensured at least one coefficient value always remained above the chance threshold (0.3) to ensure sufficient wave structure (which is essentially why three points where needed).

We compared the values of three coefficients obtained from each time instant of the working memory trials with the simulated datasets to then automatically classify the type of wave observed, based on the Euclidean distance between a wave type and the observed phase map. Specifically, the three ρ c values for each wave type could be represented as a point in three-dimensional space. The Euclidean distances between the corresponding point for an observed phase map and all wave types were computed. The phase map was classified into the wave type from which it was the closest. Coefficient matching with three different maps, allowed for greater accuracy with lesser chances of misclassification.

Choice of array points for coefficient calculation.

Our results were not constrained by the exact choice of points on the array around which the circular-circular correlation coefficient was to be calculated ( S2A Fig ). Our intuition for choosing (1,4), (4,1) and (4,4) were based on trying to capturing the whole spectrum of wave types. (1,4) was able to differentiate waves that traveled roughly towards the top-half of the array vs those that traveled towards the bottom-half ( S1B Fig ). However, (1,4) had a “chance-zone” along the horizontal axis, i.e. it could not identify waves that traveled in horizontal directions. For that purpose, another coefficient–around (4,1)—was necessary to capture waves that traveled laterally along the array. An extra point (4,4) was added to identify rotating waves that naturally occurred with their center on the array.

Inferring rotating wave locations.

In Fig 5B , the locations of the rotating wave directions were inferred based on the levels of the curvature recorded, the histogram spread observed and the number of short-wavelength waves recorded. For example, in Subject1, the right hemisphere array captured a greater incidence of short-wavelength waves–leading to the conclusion that more of the rotation is on the array—when compared to the left hemisphere. The highest wave levels were closer to the ventral side of the PFC, thus suggesting that the part of the rotation that was outside the array was more towards the ventral side. Following this logic, the rotating directions were placed on the array through a subjective estimation based on a combination of all these results.

Supporting information

(A) Quadrant-based distance map used for traveling wave identification (one quadrant shown for diagonal wave detection). The distance maps would change accordingly for other wave types. (B) Rotation maps around three points on the array (top), adjusted for circular values (mean-centered and sine-transformed; bottom) with the net directions and bisecting axes marked. (C) Example waves (white arrows), with their phase maps (top) adjusted for circular values (bottom) shown with the corresponding coefficient values on the side. (D) Plot illustrating spatial wavelengths for planar and rotating waves. (E) Plot of three correlation coefficient values for different wave types. The region between the dashed horizontal lines denotes the “chance zone”. It was ensured that at least one coefficient for each wave type remained outside this zone.

https://doi.org/10.1371/journal.pcbi.1009827.s001

(A) (left) Similar wave classification between planar and rotating waves as Fig 3 , with three different points chosen to distinguish between these wave types. The grey circles denote the points on the array around which the coefficient was calculated. (right) Plots showing the number of time instants in which rotating/planar waves were observed across all arrays and trials in the three frequency ranges. (B, C) Number of wave instants observed for different wave types across frequencies and arrays.

https://doi.org/10.1371/journal.pcbi.1009827.s002

(A) Histograms quantifying the correlation values seen on average in each trial during the sample, delay, and test-onset intervals (all 0.5 s in length), for the left dlPFC array of Animal 2. A correlation value to the right of the shaded region (positive) denotes waves in a particular direction, while the left means the opposite direction. The shaded region denotes the “chance zone” where no conclusion regarding wave direction can be made. For each frequency range (each row), the red histogram corresponds to the correlations observed in baseline conditions (0.5 s pre-fixation), while the green histogram corresponds to the correlations observed in that epoch (0.5 s). The blue dots denote if the two are significantly different from each other (p<0.01). (B) Quantification of the difference between the green and red curves in (A) for all 0.5 s intervals during each trial. The red line shows difference from baseline for the positive wave direction, while the blue line for the negative wave direction.

https://doi.org/10.1371/journal.pcbi.1009827.s003

(A) Illustration of a rotating wave with increasing wavelengths at greater distances from the center of rotation (white circle). (B) Spike rates observed for long and short wavelengths, combined across all arrays for theta (left) and beta (right) traveling waves. Star indicates statistically significant difference (p<0.01).

https://doi.org/10.1371/journal.pcbi.1009827.s004

(A) Percentage of wave vs non-wave instants for beta waves combined across arrays. Star indicates statistically significant difference (p<0.01). (B) Plot of the phase standard deviation for wave and non-wave instants (normalized probability) observed for beta waves in the left dlPFC array of Subject 2. The red peak to the left of the blue peak corresponds to the standing wave mode, indicating oscillations that were not traveling waves but had a lower phase variance (than traveling waves) on the array.

https://doi.org/10.1371/journal.pcbi.1009827.s005

S1 Movie. An example beta-band planar wave on the 8x8 left dlPFC array from a trial in Subject 2, corresponding to the panels in Fig 1 .

Each tile marks the LFP oscillation amplitude (warmer colors mean higher amplitude) at that point in time and space.

https://doi.org/10.1371/journal.pcbi.1009827.s006

S2 Movie. An example of a theta rotating wave on the 8x8 array, during the memory delay period of a trial in the right dlPFC of Subject 1.

https://doi.org/10.1371/journal.pcbi.1009827.s007

Acknowledgments

We thank Jordan G. DeFarias for his technical assistance and Jesus Ballesteros, Andre Bastos, Alex Major, Morteza Moazami, Dimitris Pinotsis, and Jefferson Roy for helpful comments.

  • View Article
  • PubMed/NCBI
  • Google Scholar

Suggestions or feedback?

MIT News | Massachusetts Institute of Technology

  • Machine learning
  • Social justice
  • Black holes
  • Classes and programs

Departments

  • Aeronautics and Astronautics
  • Brain and Cognitive Sciences
  • Architecture
  • Political Science
  • Mechanical Engineering

Centers, Labs, & Programs

  • Abdul Latif Jameel Poverty Action Lab (J-PAL)
  • Picower Institute for Learning and Memory
  • Lincoln Laboratory
  • School of Architecture + Planning
  • School of Engineering
  • School of Humanities, Arts, and Social Sciences
  • Sloan School of Management
  • School of Science
  • MIT Schwarzman College of Computing

“Traveling” nature of brain waves may help working memory work

Press contact :.

Photo of a crowd full of stadium fans in which a column of people are all standing with their arms raised

Previous image Next image

After more than a century of study, the significance of brain waves — the coordinated, rhythmic electrical activity of groups of brain cells — is still not fully known. An especially underappreciated aspect of the phenomenon is that waves spatially propagate, or “travel,” through brain regions over time. A new study by researchers at The Picower Institute for Learning and Memory at MIT measured how waves travel in the brain’s prefrontal cortex during working memory to investigate the functional advantages that this apparent motion may produce.

“Most of the neuroscience literature involves lumping electrodes together and analyzing for time variations,” like changes in power at a particular frequency, says lead author Sayak Bhattacharya, a Picower Fellow in the lab of senior author, professor of brain and cognitive sciences and Picower Institute member Earl Miller . “It is important to appreciate that there are spatial subtleties, too. Brain oscillations move across the cortex in the form of traveling waves. These waves are similar to stadium waves where nothing actually moves, but sequential on-and-off of neighbors give it the appearance of a traveling wave.”

In other words, while the neurons under an eavesdropping electrode might burst with activity at a particular frequency, it’s also true that just before they perked up, neurons nearby in some direction had already done so, and very soon some other neurons on the opposite side will follow suit. Bhattacharya, Miller, and their co-authors conducted the study , published Jan. 28 in PLOS Computational Biology, to learn what that might mean for the vital brain function of working memory, where we must hold new information in mind to put it to use. It’s how we remember the directions to the bathroom we were just told, or today’s specials at the restaurant.

To do this, they cracked open some old data they had recorded from animals while they played a simple working memory game. The animals would see a single image on a screen and after a brief pause they would then see it along with a few other images. To get a little reward, they had to fix their gaze on the original image. It’s a simple game but the little stages (look at the new image, remember it during the pause, recognize and stare at it in the group) provide distinct moments of perception, memory, and then putting them to use. By combing over electrode recordings made in the animals during sessions of the game, Bhattacharya could analyze whether the recorded waves were traveling at each moment and how.

What he found was that there were many distinct waves at various frequency bands washing back and forth across the electrodes in various directions. Careful calculations revealed that the waves were actually rotating in circle-like patterns around central anatomical points within the prefrontal cortex (again, like the wave in a football stadium rotates around the field of play). That’s notable because in other traveling wave studies usually the waves are planar, meaning they just move across from one place to another rather than going around as if on a race track.

Moreover, Miller says, the waves changed direction in particularly important ways. When the animals were idle, different directions of motion (e.g., clockwise vs. counterclockwise) were pretty much evenly mixed, but at different times during the task, specific directions became significantly more prominent in various frequency bands. This was especially true among beta frequency waves, which became much more uniform in their direction only while the animals played the game. Other frequencies became more weighted toward particular directions during specific phases of the game (like when the first image was presented). These changes suggested that the directions matter to how the brain organizes its response to the task.

“The waves are generally traveling but the brain can change how the waves travel to suit different cognitive functions,” Miller says.

Indeed there are several ways that rotating traveling waves could aid a task like working memory, he notes. For one example, a key requirement of working memory is being able to keep information at the forefront of conscious thought while it’s needed. A stationary wave (one in which all the neurons involved were “on” or “off” in unison) would mean that information could be unavailable when activity was off across the whole group. With a rotating traveling wave, there is always activity somewhere around the circle — just like how in a stadium of fans doing “the wave,” the next section stands up as soon as the preceding one sits down.

For another example, rotating waves could provide neurons with a regularly recurring stimulation with precise timing, Miller continued. That may promote strengthening connections within these coordinated groups via a phenomenon called spike-timing dependent plasticity, in which the timing of input to a neuron influences how strongly it will connect with the partner that delivered the signal. The researchers also speculate that timing might also matter in another prefrontal cortex function: making predictions.

More work needs to be done to know with certainty how traveling waves aid working memory. Bhattacharya says new insight could come from investigating how they look when working memory doesn’t work.

“This working memory task was pretty easy and our animals did them without much error,” he says. “We want to study more complicated tasks — maybe multi-item working memory — and check if the traveling waves are disrupted somehow during the error trials. This would lead to interesting insight about the computational abilities of these waves.”

In addition to Bhattacharya and Miller, the paper’s other authors are Scott Brincat and Mikael Lundqvist.

The JPB Foundation, the National Institutes of Health, and the Office of Naval Research provided funding for the study.

Share this news article on:

Related links.

  • Earl Miller
  • The Picower Institute for Learning and Memory
  • Department of Brain and Cognitive Sciences

Related Topics

  • Brain and cognitive sciences
  • Picower Institute
  • Neuroscience

Related Articles

Earl Miller presents a new model of working memory at a recent Picower Institute symposium

How working memory gets you through the day

Neuroscientists are debating the nature and meaning of neural activity during the critical moment of working memory when people must hold information in mind.

The debate over how working memory works

brain travelling waves

New study reveals how brain waves control working memory

brain travelling waves

Study: Rhythmic interactions between cortical layers underlie working memory

Previous item Next item

More MIT News

Charles Williams draws in an art class, with two students in the background.

Opening the doorway to drawing

Read full story →

Photo of the MIT Dome and surrounding Killian court in autumn with leaves on the ground

School of Engineering fourth quarter 2024 awards

Susan Solomon and Tran Hong Ha pose on a stage. They are both holding a folio open for the audience to see, and Solomon is holding a gold trophy.

Susan Solomon wins VinFuture Award for Female Innovators

Study: stars travel more slowly at milky way’s edge.

Six people pose behind a table. Two are standing, and the seated ones each have a laptop

Entrepreneur creates career pathways with MIT OpenCourseWare

A male student clad in a waterproof rain suit covered in multicolored paint stands onstage between two Blue Man Group members with his arms raised, before an audience of people wearing clear ponchos.

Performance art and science collide as students experience “Blue Man Group”

  • More news on MIT News homepage →

Massachusetts Institute of Technology 77 Massachusetts Avenue, Cambridge, MA, USA

  • Map (opens in new window)
  • Events (opens in new window)
  • People (opens in new window)
  • Careers (opens in new window)
  • Accessibility
  • Social Media Hub
  • MIT on Facebook
  • MIT on YouTube
  • MIT on Instagram

brain travelling waves

‘Traveling’ nature of brain waves may help working memory work

The act of holding information in mind is accompanied by coordination of rotating brain waves in the prefrontal cortex, a new study finds.

After more than a century of study, the significance of brain waves – the coordinated, rhythmic electrical activity of groups of brain cells – is still not fully known. An especially underappreciated aspect of the phenomenon is that waves spatially propagate, or “travel,” through brain regions over time. A new study by researchers at The Picower Institute for Learning and Memory at MIT measured how waves travel in the brain’s prefrontal cortex during working memory to investigate the functional advantages that this apparent motion may produce.

“Most of the neuroscience literature involves lumping electrodes together and analyzing for time variations,” like changes in power at a particular frequency, said lead author Sayak Bhattacharya, a postdoctoral Picower Fellow in lab of senior author and Picower Professor Earl Miller . “It is important to appreciate that there are spatial subtleties, too. Brain oscillations move across the cortex in the form of traveling waves. These waves are similar to stadium waves where nothing actually moves, but sequential on-and-off of neighbors give it the appearance of a traveling wave.”

Above: A stadium wave forms when fans in adjacent sections stand up and then sit back down in  sequence around the seating area. This creates a wave that travels through the crowd even though no individual fans leave their seats. A new study finds that working memory is accompanied by brain waves rotating around central points, analogous to this action.  Image by Ken Lund from Reno, Nevada, USA, CC BY-SA 2.0 . via Wikimedia Commons.

In other words, while the neurons under an eavesdropping electrode might burst with activity at a particular frequency, it’s also true that just before they perked up, neurons nearby in some direction had done so and very soon some other neurons on the opposite side will follow suit. Bhattacharya, Miller and their co-authors conducted the study published in PLOS Computational Biology to learn what that might mean for the vital brain function of working memory, where we must hold new information in mind to put it to use. It’s how we remember the directions to the bathroom we were just told, or today’s specials at the restaurant.

To do this, they cracked open some old data they had recorded from animals while they played a simple working memory game. The animals would see a single image on a screen and after a brief pause they would then see it along with a few other images. To get a little reward, they had to fix their gaze on the original image. It’s a simple game but the little stages (look at the new image, remember it during the pause, recognize and stare at it in the group) provide distinct moments of perception, memory and then putting them to use. By combing over electrode recordings made in the animals during sessions of the game, Bhattacharya could analyze whether the recorded waves were traveling at each moment and how.

What he found was that there were many distinct waves at various frequency bands washing back and forth across the electrodes in various directions. Careful calculations revealed that the waves were actually rotating in circle-like patterns around central anatomical points within the prefrontal cortex (again, like the wave in a football stadium rotates around the field of play). That’s notable because in other traveling wave studies usually the waves are planar, meaning they just move across from one place to another rather than going around as if on a race track.

Moreover, Miller said, the waves changed direction in particularly important ways. When the animals were idle, different directions of motion (e.g. clockwise vs. counterclockwise) were pretty much evenly mixed but at different times during the task, specific directions became significantly more prominent in various frequency bands. This was especially true among beta frequency waves, which became much more uniform in their direction only while the animals played the game. Other frequencies became more weighted toward particular directions during specific phases of the game (like when the first image was presented). These changes suggested that the directions matter to how the brain organizes its response to the task.

“The waves are generally traveling but the brain can change how the waves travel to suit different cognitive functions,” Miller said.

Indeed there are several ways that rotating traveling waves could aid a task like working memory, he noted. For one example, a key requirement of working memory is being able to keep information at the forefront of conscious thought while it’s needed. A stationary wave (one in which all the neurons involved were “on” or “off” in unison) would mean that information could be unavailable when activity was off across the whole group. With a rotating traveling wave there is always activity somewhere around the circle – just like how in a stadium of fans doing “the wave,” the next section stands up as soon as the preceding one sits down.

For another example, rotating waves could provide neurons with a regularly recurring stimulation with precise timing, Miller continued. That may promote strengthening connections within these coordinated groups via a phenomenon called spike-timing dependent plasticity in which the timing of input to a neuron influences how strongly it will connect with the partner that delivered the signal. The researchers also speculate that timing might also matter in another prefrontal cortex function: making predictions.

More work needs to be done to know with certainty how traveling waves aid working memory. Bhattacharya said new insight could come from investigating how they look when working memory doesn’t work.

“This working memory task was pretty easy and our animals did them without much error,” he said. “We want to study more complicated tasks—maybe multi-item working memory—and check if the traveling waves are disrupted somehow during the error trials. This would lead to interesting insight about the computational abilities of these waves.”

In addition to Bhattacharya and Miller, the paper’s other authors are Scott Brincat and Mikael Lundqvist.

The JPB Foundation, the National Institutes of Health and the Office of Naval Research provided funding for the study.

Related Articles

Study reveals a universal pattern of brain wave frequencies.

brain travelling waves

Anesthesia blocks sensation by cutting off communication within the cortex

A blue-hued cartoon shows a transparent head on the left in profile with a brain inside. Big slow waves emanate from marked points in the brain into the space on the right.

Anesthesia technology precisely controls unconsciousness in animal tests

An operating room scene shows a patient on a table. Our perspective is from behind the anesthesiologist who holds a mask on the patient's face and watches a monitor with a bunch of indicators. A surgeon stands out of focus on the far end of the patient.

A multifunctional tool for cognitive neuroscience

Indie garwood, wearing latex gloves and a lab coat, streches out a fiber in the lab.

Traveling and standing waves in the brain

Affiliation.

  • 1 Section on Functional Imaging Methods, National Institute of Mental Health, National Institutes of Health, Bethesda, MD, USA. [email protected].
  • PMID: 35902650
  • PMCID: PMC10170397
  • DOI: 10.1038/s41593-022-01119-0

Studying the natural wanderings of the living brain is extremely challenging. Bolt et al. describe a new framework to consider the brain’s intrinsic activity based on the geophysical concepts of standing and traveling waves.

Publication types

  • Research Support, N.I.H., Intramural
  • Brain Waves*

Grants and funding

  • Z99 MH999999/ImNIH/Intramural NIH HHS/United States
  • ZIA MH002783/ImNIH/Intramural NIH HHS/United States

salk institute logo

October 7, 2020

Traveling brain waves help detect hard-to-see objects

Salk scientists discover patterns of neural waves in the awake brain that help detect objects

Home - Salk News - Traveling brain waves help detect hard-to-see objects

LA JOLLA—Imagine that you’re late for work and desperately searching for your car keys. You’ve looked all over the house but cannot seem to find them anywhere. All of a sudden you realize your keys have been sitting right in front of you the entire time. Why didn’t you see them until now?

Now, a team of Salk Institute scientists led by Professor John Reynolds has uncovered details of the neural mechanisms underlying the perception of objects. They found that patterns of neural signals, called traveling brain waves, exist in the visual system of the awake brain and are organized to allow the brain to perceive objects that are faint or otherwise difficult to see. The findings were published in Nature on October 7, 2020.

Top from left: Zac Davis and Terrence Sejnowski. Bottom from left: Lyle Muller and John Reynolds.

Click here for a high-resolution image.

“We’ve discovered that faint objects are much more likely to be seen if visualizing the object is timed with the traveling brain waves. The waves actually facilitate perceptual sensitivity, so there are moments in time when you can see things that you otherwise could not,” says Reynolds, senior author of the paper and holder of the Fiona and Sanjay Jha Chair in Neuroscience. “It turns out that these traveling brain waves are an information-gathering process leading to the perception of an object.”

Scientists have studied traveling brain waves during anesthesia but dismissed the waves as an artifact of the anesthesia. Reynolds’ team, however, wondered if these waves exist in the visual part of the brain while awake and if they play a role in perception. They combined recordings in the visual cortex with cutting-edge computational techniques that enabled them to detect and track traveling brain waves.

“In order to understand the neural mechanisms of perception, we needed to develop new computational techniques to track neuronal activity in the visual cortex moment by moment,” says co-first author Lyle Muller, BrainsCAN-funded assistant professor in the Department of Applied Mathematics and the Brain and Mind Institute at Western University in Ontario, Canada, and previously a postdoctoral fellow in the Sejnowski lab at Salk. “We then used these computational methods to uncover what change was occurring in the nervous system to suddenly allow for object recognition.”

The scientists recorded the activity of the neurons from an area of the brain that contained a complete map of the visual world. They then tracked the trajectories of the traveling brain waves during a visual perception task. The scientists held an onscreen target at the threshold of visibility, so that observers could only detect the object 50 percent of the time, and recorded when the target was spotted. Since the target was not changing, the researchers reasoned that the observer’s ability to perceive the object only half of the time had to be due to some change in the neural signals inside the brain.

They found that the brain’s ability to recognize targets was directly related to when and where the traveling brain waves occurred in the visual system: when the traveling waves aligned with the stimulus, the observer could detect the target more easily. These traveling brain waves, which occurred several times per second, were similar to a stadium of sports fans successively standing up and raising their arms, then lowering them and sitting down again. It appears that the visual system is actively sensing the external environment, according to the team.

“There is a spontaneous level of activity in the brain that appears to be regulated by these traveling waves,” says Salk Professor Terrence Sejnowski , an author of the paper and holder of the Francis Crick Chair. “We think the waves are the product of the activity that is propagating around the brain, driven by local neurons firing.”

“We go about our everyday lives thinking that we are accurately seeing the world, but, in fact, our brains are filling in details that are difficult to see,” says Zac Davis, co-first and corresponding author of the paper and a Salk postdoctoral fellow in the Reynolds lab. “Now, we have discovered how the brain weaves together hard-to-see information to perceive an object.”

In the future, the scientists plan to examine whether these brain waves are coordinated across different brain regions devoted to vision. The researchers theorize that the brain waves could serve as a gate between the sensory processing and conscious perception that emerges from the brain as a whole.

Julio-Martinez Trujillo of Western University was also an author on this paper.

The work was supported by the Dan and Martina Lewis Biophotonics Fellowship; the Gatsby Charitable Foundation; the Fiona and Sanjay Jha Chair in Neuroscience; the Canadian Institute for Health Research; the Swartz Foundation; and the National Institutes of Health (R01-EY028723, T32 EY020503-06, and T32 MH020002-16A).

DOI: 10.1038/s41586-020-2802-y

PUBLICATION INFORMATION

Spontaneous Traveling Cortical Waves Gate Perception in Behaving Primates

Zachary W. Davis, Lyle Muller, Julio-Martinez Trujillo, Terrence Sejnowski, and John H. Reynolds

Neuroscience

Neuroscience

For more information.

Office of Communications Tel: (858) 453-4100 [email protected]

The Salk Institute For Biological Studies:

Unlocking the secrets of life itself is the driving force behind the Salk Institute. Our team of world-class, award-winning scientists pushes the boundaries of knowledge in areas such as neuroscience, cancer research, aging, immunobiology, plant biology, computational biology and more. Founded by Jonas Salk, developer of the first safe and effective polio vaccine, the Institute is an independent, nonprofit research organization and architectural landmark: small by choice, intimate by nature, and fearless in the face of any challenge.

Connect With Us

Keep in touch.

Inside Salk

Explore our magazine.

Every cure begins with you.

Latest discoveries, events & more.

  • Phone This field is for validation purposes and should be left unchanged.

brain travelling waves

Your partnership allows our scientists to accelerate the pace of high-risk, high-reward discoveries that have the potential to benefit the health of all humanity-be it cancer, Alzheimer’s disease, climate change, infectious diseases or more.

Donate to Salk Institute today

  • The Magazine
  • Stay Curious
  • The Sciences
  • Environment
  • Planet Earth

When Brain Waves Go Traveling

Neuroskeptic icon

In July last year I asked, Could Traveling Waves Upset Cognitive Neuroscience? This was a post about a paper from David Alexander et al. arguing that neuroscience was overlooking the importance of how neural activity moves or travels through the brain. Now Alexander et al. are back with a new PLoS ONE paper in which they describe traveling waves in human brain activity, as measured with magnetoencephalography ( MEG ). The authors scanned 20 volunteers during a visual and auditory task. Alexander et al. focussed on "the class of waves which are characterized by a linear trajectory in the Cartesian coordinates of the sensor space." Here's a visualization. On the left we see a classical standing wave, on the right is a traveling wave. On this image, the horizontal axis is time and the vertical axis is space - each row represents one of the MEG scanner's 151 sensors, arranged around the volunteers' head. The color represents the phase of the wave, in this case in the alpha band with a frequency of 9.2 Hz. The diagonal stripe in the right panel shows a wave moving in space. A movie representation of a traveling wave can be seen here .

So what were these traveling waves? They tended to have a long spatial wavelength, mostly 30-35 cm - which means that they spanned pretty much the entire head. Alexander et al. say that there may also be more localized waves (shorter spatial wavelengths) which MEG cannot detect, because the skull filters them out. The speed of the waves ( group velocity ) was correlated with their frequency. For the fastest beta waves (28 Hz), the measured speed was around 4 meters per second - and the true speed, in the brain, will have been higher, because the brains folds mean that the path distances are longer. The authors conclude

Overall, wave events have broad distributions of both trajectories and speeds. The waves are ubiquitous, with clear episodes across the frequency range, several times a second...

What's more, Alexander et al. suggest that conventional analysis of MEG (and EEG ) data, which considers standing waves and ignores traveling ones, is missing much of the picture. They give as an example in which the average response to a stimulus (left) appears to be a standing wave, but cluster analysis of the individual trials reveals some travelling waves (e.g. B1) as well as different standing waves (e.g. B6).

So what's going on? I asked Alexander

"Do we know how these brain-wide waves happen? Are they purely cortical in origin, or are they co-ordinated by input e.g. from the thalamus ?"

He replied:

we don’t “know” whether they are purely cortical in origin. They clearly are heavily cortical, but there’s not the evidence yet. I could only give a bunch of speculations, and this would mainly reflect my own bias that connection topology is more important than the evolutionary accretion of anatomical structures.

Alexander DM, Nikolaev AR, Jurica P, Zvyagintsev M, Mathiak K, & van Leeuwen C (2016). Global Neuromagnetic Cortical Fields Have Non-Zero Velocity. PloS ONE, 11 (3) PMID: 26953886

Already a subscriber?

Register or Log In

Discover Magazine Logo

Keep reading for as low as $1.99!

Sign up for our weekly science updates.

Save up to 40% off the cover price when you subscribe to Discover magazine.

Facebook

U.S. flag

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

  • Publications
  • Account settings
  • Advanced Search
  • Journal List

Logo of elife

DMT alters cortical travelling waves

Andrea alamia.

1 Cerco, CNRS Université de Toulouse, Toulouse, France

Christopher Timmermann

2 Computational, Cognitive and Clinical Neuroscience Laboratory (C3NL), Faculty of Medicine, Imperial College, London, United Kingdom

3 Centre for Psychedelic Research, Division of Psychiatry, Department of Brain Sciences, Imperial College London, London, United Kingdom

David J Nutt

Rufin vanrullen.

4 Artificial and Natural Intelligence Toulouse Institute (ANITI), Toulouse, France

Robin L Carhart-Harris

Associated data.

Alamia A. 2020. DMT alters cortical travelling waves. Open Science Framework. [ CrossRef ]

The data and the code to perform the analysis are available at : https://osf.io/wujgp/ .

The following dataset was generated:

Psychedelic drugs are potent modulators of conscious states and therefore powerful tools for investigating their neurobiology. N,N, Dimethyltryptamine (DMT) can rapidly induce an extremely immersive state of consciousness characterized by vivid and elaborate visual imagery. Here, we investigated the electrophysiological correlates of the DMT-induced altered state from a pool of participants receiving DMT and (separately) placebo (saline) while instructed to keep their eyes closed. Consistent with our hypotheses, results revealed a spatio-temporal pattern of cortical activation (i.e. travelling waves) similar to that elicited by visual stimulation. Moreover, the typical top-down alpha-band rhythms of closed-eyes rest were significantly decreased, while the bottom-up forward wave was significantly increased. These results support a recent model proposing that psychedelics reduce the ‘precision-weighting of priors’, thus altering the balance of top-down versus bottom-up information passing. The robust hypothesis-confirming nature of these findings imply the discovery of an important mechanistic principle underpinning psychedelic-induced altered states.

Introduction

N,N, Dimethyltryptamine (DMT) is a mixed serotonin receptor agonist that occurs endogenously in several organisms ( Christian et al., 1977 ; Nichols, 2016 ) including humans ( Smythies et al., 1979 ), albeit in trace concentrations. DMT, which is a classic psychedelic drug, is also taken exogenously by humans to alter the quality of their consciousness. For example, synthesized compound is smoked or injected but it has also been used more traditionally in ceremonial contexts (e.g. in Amerindian rituals). When ingested orally, DMT is metabolized in the gastrointestinal (GI) system before reaching the brain. Its consumption has most traditionally occurred via drinking ‘Ayahuasca’, a brew composed of plant-based DMT and β –carbolines (monoamine oxidize inhibitors), which inhibit the GI breakdown of the DMT ( Buckholtz and Boggan, 1977 ). Modern scientific research has mostly focused on intravenously injected DMT. Administered in this way, DMT’s subjective effects have a rapid onset, reaching peak intensity after about 2–5 min and subsiding thereafter, with negligible effects felt after about 30 min ( Strassman, 2001 ; Strassman, 1995a ; Timmermann et al., 2019 ).

Previous electrophysiological studies investigating changes in spontaneous (resting state) brain function elicited by ayahuasca have reported consistent broadband decreases in oscillatory power ( Riba et al., 2002 ; Timmermann et al., 2019 ), while others have noted that the most marked decreases occur in α-band oscillations (8–12 Hz) ( Schenberg et al., 2015 ). Alpha decreases correlated inversely with the intensity of ayahuasca-induced visual hallucinations ( Valle et al., 2016 ) and are arguably the most reliable neurophysiological signature of the psychedelic state identified to-date ( Muthukumaraswamy et al., 2013 ) – with increased signal diversity or entropy being another particularly reliable biomarker ( Schartner et al., 2017 ). In the first EEG study of the effects of pure DMT on on-going brain activity, marked decrease in the α and β (13–30 Hz) band power was observed as well as increase in signal diversity ( Timmermann et al., 2019 ). Increase in lower frequency band power (δ = 0.5–4 Hz and θ = 4–7 Hz) also became evident when the signal was decomposed into its oscillatory component. Decreased alpha power and increased signal diversity correlated most strongly with DMT’s subjective effects – consolidating the view that these are principal signatures of the DMT state, if not the psychedelic state more broadly.

Focusing attention onto normal brain function, outside of the context of psychoactive drugs, electrophysiological recordings in cortical regions reveal distinct spatio-temporal dynamics during visual perception, which differ considerably from those observed during closed-eyes restfulness. It is possible to describe these dynamics as oscillatory ‘travelling waves’, i.e. fronts of rhythmic activity which propagate across regions in the cortical visual hierarchy ( Lozano-Soldevilla and VanRullen, 2019 ; Muller et al., 2014 ; Sato et al., 2012 ). Recent results showed that travelling waves can spread from occipital to frontal regions during visual perception, reflecting the forward bottom-up flow of information from lower to higher regions. Conversely, top-down propagation from higher to lower regions appears to predominate during quiet restfulness ( Alamia and VanRullen, 2019 ; Halgren et al., 2019 ; Pang et al., 2020 ).

Taken together these results compel us to ask how travelling waves may be affected by DMT, particularly given their association with predictive coding ( Alamia and VanRullen, 2019 ; Friston, 2019 ) and a recent predictive coding inspired hypothesis on the action of psychedelics (‘REBUS’) – which posits decreased top-down processing and increased bottom-up signal passing under these compounds ( Carhart-Harris and Friston, 2019 ). Moreover, DMT lends itself particularly well to the testing of this hypothesis as its visual effects are so pronounced. Given that visual perception is associated with an increasing in forward travelling waves and eyes-closed visual imagery under DMT can feel as if one is ‘seeing with eyes shut’ ( de Araujo et al., 2012 ) – does a consistent increase in forward travelling waves under DMT account for this phenomenon?

Here we sought to address these questions by quantifying the amount and direction of travelling waves in a sample of healthy participants who received DMT intravenously, during eyes-closed conditions. We hypothesized that DMT acts by disrupting the normal physiological balance between top-down and bottom-up information flow, in favour of the latter ( Carhart-Harris and Friston, 2019 ). Moreover, we ask: does this effect correlate with the vivid ‘visionary’ component of the DMT experience? Providing evidence in favour of this hypothesis would indicate that forward travelling waves do play a crucial role in conscious visual experience, irrespective of the presence of actual photic stimulation.

Quantifying travelling waves

As demonstrated by both theoretical and experimental evidence ( Nunez, 2000 ; Nunez and Srinivasan, 2014 ; Nunez and Srinivasan, 2009 ), in most systems, including the human brain, travelling waves occur in groups (or packets) over some range of spatial wavelengths having multiple spatial and temporal frequencies. Given any configurations of electrodes, only parts of these packets can be successfully detected, i.e. waves shorter than the spatial extent of the array, and waves longer than twice the electrode separation distance (Nyquist criterion in space). In scalp recordings, the shorter waves may be mostly removed by volume conduction. As a consequence, waves recorded directly from the cortex emphasize shorter waves than the scalp recorded waves. Specifically, in the case of small cortical arrays, the overlap between cortical and scalp data may be minimal, and the estimated wave properties (including propagation direction) may differ. Additionally, it is important to consider that when waves are travelling in multiple directions at nearly the same time in ‘closed’ systems (e.g. the cortical/white matter), waves either damp out or interfere with each other to form standing waves (e.g. alpha waves travelling both forward and backward). It is reasonable to assume that the behaviour of these properties will relate to global brain and mind states, and be sensitive to state-altering psychoactive drugs ( Nunez, 2000 ; Nunez and Srinivasan, 2014 ; Nunez and Srinivasan, 2009 ).

Practically, we measure the waves’ amount and direction with a method devised in our previous studies ( Alamia and VanRullen, 2019 ; Pang et al., 2020 ). We slide a one-second time-window over the EEG signals (with 0.5 s overlap). For each time-window, we generate a 2D map (time/electrodes) by stacking the signals from five central mid-line electrodes (Oz to FCz, see Figure 1 ). For each map, we then compute a 2D-FFT, in which the upper- and lower-left quadrant represent the power of forward (FW) and backward (BW) travelling waves, respectively (since the 2D-FFT is symmetrical around the origin, the lower- and upper-right quadrants contain the same information). From both quadrants we extracted the maximum values, representing the raw amount of FW and BW waves in that time-window. Next, we performed the same procedure after having shuffled the electrodes’ order, thereby disrupting spatial information (including the waves’ directionality) while retaining the same overall spectral power. In other words, the surrogate measures reflect the amount of waves expected solely due to the temporal fluctuations of the signal. After having computed the maximum values for the FW and BW waves of the surrogate 2D-FFT spectra one hundred times (and averaging the 100 values), we compute the net amount of FW and BW waves in decibel (dB), by applying the following formula:

An external file that holds a picture, illustration, etc.
Object name is elife-59784-fig1.jpg

From each 1 s EEG epoch we extract a 2D-map, obtained by stacking signals from five midline electrodes. For each map we compute a 2D-FFT in which the maximum values in the upper- and lower-left quadrants represent respectively the amount of forward (FW – in blue) and backward (BW – in red) waves. For each map, we also compute surrogate values by shuffling the electrodes’ order 100 times, so as to retain temporal fluctuations while disrupting the spatial structure of the signals (including any travelling waves). Eventually, we compute the wave strength in decibel (dB) by combining the real and the surrogate values.

where W represents the maximum value extracted for each quadrant (i.e. forward FW or backward BW), and Wss the respective surrogate value. Importantly, this value – expressed in decibel – represents the net amount of waves against the null distribution. In other words, it is informative to compare this value to zero, to assess the significance of waves. On the other hand, a direct comparison between FW and BW waves in each time-bin is not readily interpretable, as it is possible to simultaneously record waves propagating in both directions—as observed during visual stimulation epochs (see below). In addition, it’s important to note that our waves’ analysis focuses on the sensor level, as source projections presents a number of important limitations, such as impairing long-range connections, as well as smearing of signals due to scalp interference ( Alexander et al., 2019 ; Freeman and Barrie, 2000 ; Nunez, 1974 ).

Does DMT influence travelling waves?

After defining our measure of the waves’ amount and direction, we investigated whether the intake of DMT alters the cortical pattern of travelling waves. Participants underwent two sessions in which they were injected with either placebo or DMT (see Materials and methods for details). Importantly, during all of the experiments, participants rested in a semi-supine position, with their eyes closed. EEG recordings were collected 5 min prior to drug administration and up to 20 min after. The left column of Figure 2A shows the amount of BW and FW waves in the 5 min preceding and following drug injection (either placebo or DMT). Consistent with previous observations on independent data ( Alamia and VanRullen, 2019 ), during quiet closed-eyes restfulness a significant amount of BW waves spread from higher to lower regions (as confirmed by a Bayesian t-test against zero for both DMT and Placebo conditions, BFs 10  >>100, error <0.01%, 95% Credible Intervals (CI) DMT: [0.221, 0.637], Placebo: [0.273, 0.666]), whereas no significant waves propagate in the opposite FW direction (Bayesian t-test against zero: BFs 10  <0.15, error <0.01%; 95% CI DMT: [−0.424, 0.088], Placebo: [−0.372, 0.110]). However, after DMT injection, the cortical pattern changed drastically: the amount of BW waves decreased (but remaining significantly above zero – BFs 10  = 12.6, 95% CI: [0.057, 0.322]), whereas the amount of FW waves increased significantly above zero (BF 10  = 5.4 95% CI: [0.027, 0.336]). These results, obtained by comparing the amount of waves before and after injection (pre-post factor) of Placebo or DMT (drug factor), were confirmed by two Bayesian ANOVA performed separately on BW and FW waves (all factors including interactions reported BFs 10  >>100, error <2%), and were not confounded by differences in dosage (see Materials and methods and Figure 2—figure supplement 1 ). A power analysis comparing DMT and Placebo conditions after infusion for both FW and BW direction revealed values above 90% (FW case: μ DMT =0.19, μ PLACEBO = -0.20 and σ = 0.29 yields to power equals to 0.9168; BW case: μ DMT = 0.18, μ PLACEBO = 0.51 and σ = 0.25 gives power equals to 0.9205; in both cases, we considered a type I error rate of 5%).

An external file that holds a picture, illustration, etc.
Object name is elife-59784-fig2.jpg

( A ) In the left panels the net amount of FW (blue, upper panel) and BW (red, lower panel) waves is represented pre- and post-DMT infusion. While BW waves are always present, FW waves only rise significantly above zero after DMT injection, despite participants having closed eyes. Asterisks denote values significantly different than zero, or between conditions. The panels to the right describe the minute-by-minute changes in the net amount of waves. Asterisks denote FDR-corrected p-values for amount of waves significantly different than zero. ( B ) Comparison between the waves’ temporal evolution after DMT injection (left panel) and with or without visual stimulation (right panel, from a different experiment in which participants, with open eyes, either watched a visual stimulus or a blank screen  Pang et al., 2020 ). Remarkably, the waves’ temporal profiles are very similar in the two conditions, for both FW and BW. ( C ) Comparison between changes in absolute power (as extracted from the 2D-FFT, that is FW and BW in Figure 1 ) due to DMT, placebo and visual stimulation. Remarkably, true photic visual stimulation and eyes-closed DMT induce comparably large reductions in absolute power. In fact, the effect with DMT appears to be even more pronounced (formal contrast not appropriate). Note that in the previous panels the changes in the net amount of waves were reported in dB, and occurred irrespective of the global power changes measured in panel C.

Figure 2—figure supplement 1.

An external file that holds a picture, illustration, etc.
Object name is elife-59784-fig2-figsupp1.jpg

Each line represents a different subject, whereas mean ± standard deviations are represented for each dosage, pre/post infusion for BW (red) and FW (blue) waves. Irrespective of the dosage, the amount of BW waves decreased after DMT infusion, whereas FW waves increased consistently for each subject.

Figure 2—figure supplement 2.

An external file that holds a picture, illustration, etc.
Object name is elife-59784-fig2-figsupp2.jpg

The difference in the pre-post DMT infusion observed along the sagittal line of electrodes (i.e. the one chosen for the first analysis, as reported in Figure 2A of the manuscript) is replicated considering another series of electrodes running from occipital to frontal regions between hemisphere, specifically from electrode P4 to F3 in the FW direction (diag1, Bayesian t-test BF = 4.059, error = 0.002%), and from P3 to F4 (diag2, Bayesian t-test BF = 4.848, error = 0.0001%). Interestingly, DMT induces a similar increase in FW waves, but less of a decrease in the BW direction (diag1 BW: BF = 1.948, error = 0.006%; diag1 BW: BF = 1.567, error = 0.002%). We also investigated a coronal line of electrodes, revealing waves travelling in a leftward and rightward direction above chance level (i.e. larger than 0 dB), but in-line with our hypothesis this pattern was not altered by DMT infusion (for both leftward and rightward waves BF <0.4, error ~0.02%). The bottom panel shows changes in absolute power (as extracted from the 2D-FFT, i.e. FW and BW in Figure 1 ) in each lines of electrodes. Due to DMT, we observed overall a large reduction in absolute power, in-line with previous results.

In order to explore different propagation axes than the midline, we ran the same analysis on one array of electrodes running from posterior right to anterior left regions, and one from posterior left to anterior right ones: in both cases we obtained similar results as for the midline electrodes, i.e. an increase and a decrease of FW and BW waves, respectively, following DMT infusion (see Figure 2—figure supplement 2 ). This suggests that the dominant natural propagation spread of travelling waves is along the axis that connects the furthest posterior and frontal recording channels. As a control, we additionally demonstrated that waves propagating from leftward to rightward regions (and vice versa), were not affected by DMT (see Figure 2—figure supplement 2 ). Besides, in-line with previous work on travelling waves ( Alexander et al., 2013 ; Alexander et al., 2006 ), an additional analysis based on relative phases of the alpha band-pass signals over all channels, confirmed the same results, with DMT indeed disrupting the typical top-down propagation of alpha-band waves. Furthermore, we ran a more temporally precise analysis, on a minute-by-minute scale, testing the amount of FW and BW waves in the two conditions, as shown in the right panels of Figure 2A . in-line with previous studies ( Strassman, 1995a ; Strassman, 1994 ; Timmermann et al., 2019 ), the changes in cortical dynamics appeared rapidly after intravenous DMT injection, and began to fade after about 10 min. Confirming our previous analysis, we observed an increase in FW waves (asterisks in the upper-right panel of Figure 2A show FDR-corrected significant p-values when testing against zero) and a decrease in BW waves, which, nonetheless, remained above zero (all FDR-corrected p-values<0.05). To our initial surprise, the dynamics elicited by DMT injection were remarkably reminiscent of those observed in another study, in which healthy participants alternated visual stimulation with periods of blank screen, without any drug manipulation ( Pang et al., 2020 ). Although a direct comparison is not statistically possible (because the two studies involved distinct subject groups and different EEG recording setups), we indirectly investigated the similarities between these two scenarios.

Comparison with perceptual stimulation

We recently showed that FW travelling waves increase during visual stimulation, whereas BW waves decrease, in-line with their putative functional role in information transmission ( Pang et al., 2020 ). In Figure 2B , for the sake of comparison, we contrast the cortical dynamics induced by DMT (left panel) with the results of our previous study (right panel Pang et al., 2020 ), in which participants perceived a visual stimulus (label ‘ON’) or stared at a dark screen (label ‘OFF’). Remarkably, mutatis mutandis, both FW and BW waves share a similar profile across the two conditions, increasing and decreasing respectively following DMT injection or visual stimulation. If we consider the absolute (maximum) power values derived from the 2D-FFT of each map (i.e. before estimating the surrogates and the waves’ net amount in decibel) as an estimate of spectral power, we can read the results reported in Figure 2C as an overall decrease in oscillatory power following DMT injection, more specifically in the frequency band with the highest power values (i.e. alpha band, but see next paragraph) ( Muthukumaraswamy et al., 2013 ; Riba et al., 2002 ; Schenberg et al., 2015 ; Timmermann et al., 2019 ). Such decrease in oscillatory power is also matched by a similar decrease induced by visual stimulation (all Bayesian t-test BFs 10  >>100). These results demonstrate that, despite participants having their eyes-closed throughout, DMT produces spatio-temporal dynamics similar to those elicited by true visual stimulation. These results therefore shed light on the neural mechanisms involved in DMT-induced visionary phenomena.

Does DMT influence the frequency of travelling waves?

Previous studies showed that DMT alters specific frequency bands (e.g. alpha-band Schenberg et al., 2015 ), mostly by decreasing overall oscillatory power ( Riba et al., 2002 ; Timmermann et al., 2019 ). Here, we investigated whether DMT influences not only the waves’ direction but also their frequency spectrum. We compared the frequencies of the maximum peaks extracted from the 2D-FFT (see Figure 1 ) before and after DMT or Placebo injection. Before infusion, both FW and BW waves had a strong alpha-range oscillatory rhythm ( Figure 3A , labeled ‘PRE’). Remarkably, following DMT injection, the waves’ spectrum changed drastically, with a significant reduction in the alpha-band, coupled with an increase in the delta and theta bands, for both FW (δ-band: BF 10  = 391.16, θ-band: BF 10  = 19.23, α-band: BF 10  = 16.04, β-band: BF 10  = 0.64; all errors < 0.001%) and BW waves (δ-band: BF 10  = 82.56, θ-band: BF 10  = 30.58, α-band: BF 10  = 549.54, β-band: BF 10  = 1.43; all errors < 0.005%). This result corroborates a previous analysis performed on EEG recordings from the same dataset ( Timmermann et al., 2019 ) as well as independent data pertaining to O-Phosphoryl-4-hydroxy-N,N-DMT (psilocybin), a related compound ( Muthukumaraswamy et al., 2013 ). Moreover, we investigated how DMT influences the amount of waves at each frequency.

An external file that holds a picture, illustration, etc.
Object name is elife-59784-fig3.jpg

( A ) Left and right panels show the waves’ frequencies computed from the maximum value from each quadrant in the 2D-FFT map for FW and BW waves, pre- and post-infusion. The histogram reflects the average between participants of the number of 1 s time-windows having a wave peak at the corresponding frequency. Notably, DMT significantly reduces α and β band oscillations, while enhancing δ and θ. Asterisks denote significant differences between DMT and Placebo conditions. ( B ) The upper panels show the amount of waves computed at each frequency of the 2D-FFT (i.e. not considering the maximum power per quadrant as in ( A ), but considering it for each frequency), for FW and BW waves, pre- and post-infusion. As shown in previous analysis, DMT induces an overall decrease of spectral power, especially in the alpha band BW waves, with the notable exception of an increase in FW waves in the alpha range.

As shown in Figure 3B , and in agreement with previous analyses, DMT induces an overall reduction in the amount of waves at each frequency, specifically in the alpha-band BW waves, but with the notable exception in the FW alpha band, in which DMT induces an increase in the waves’ direction.

What’s the relationship between FW and BW waves?

From the left panel of Figure 2B , it seems that during the first minutes after DMT injection, both FW and BW waves are simultaneously present in the brain. In an attempt to understand the overall relationship between FW and BW waves, we focused on the minutes when both BW and FW waves were significantly larger than 0 (minutes 2 to 5 after DMT injection, see Figure 2A ). On these data we performed a moment-by-moment correlation between their respective net amount (as measured in decibel – see Figure 1 ). We found a clear and significant negative relationship (Bayesian t-test against zero, pre-DMT BF 10  = 393.1, error <0.0001%, 95% CI: [−0.448,–0.212]; Post-DMT BF 10  = 381.9, error <0.0001%, 95% CI: [−0.479,–0.226]), very consistent across participants and irrespective of DMT injection (difference between pre- and post-, Bayesian t-test BF 10  = 0.225; error<0.02% Figure 4 , first panel). This result demonstrates that, in general, FW waves tend to be weaker whenever BW waves are stronger, and vice versa. In other words, FW and BW remain present after drug injection, sum to a consistent total amount, and remain inversely related; it is only the ratio of contribution from each that changes after DMT (i.e. less BW, more FW waves).

An external file that holds a picture, illustration, etc.
Object name is elife-59784-fig4.jpg

There is a negative correlation between the net amount of FW and BW waves, which is not influenced by the ingestion of DMT (left panel). The middle and the right panel show the relationship for a typical subject pre- and post-DMT injection.

Is there a correlation between waves and subjective reports?

We investigated whether changes in travelling waves under DMT correlated with the subjective effects of the drug. Specifically, for 20 min after DMT injection participants provided an intensity rating every minute and, when subjective effects faded, participants filled various questionnaires that addressed different aspects of the experience (see Timmermann et al., 2019 for details). First, we found a robust correlation between minute-by-minute intensity rates and the amplitude of the waves, as shown in the first panel of Figure 5 . This result reveals that the developing intensity of the drug’s subjective effects and changes in the amplitude of waves correlate positively (FW) or negatively (BW) across time, both peaking a few minutes after drug injection. Second, treating each time point independently, we again correlated intensity ratings with the amount of each wave type, across subjects. The middle panel of Figure 5 shows a clear trend for the correlation coefficients over time. Despite the limited number of data-points (n = 12), the correlation coefficients reach high values (~0.4), implying that, around the moment where the drug had its maximal effect (2–5 min after injection), those subjects who reported the most intense effects were also those who had the strongest travelling waves in the FW direction, and the weakest waves in the BW direction. Finally, we correlated the amount of FW and BW waves with ratings focused specifically on visual imagery: remarkably, ratings of all of the relevant questionnaire items correlated strongly with the increased amount of FW waves under DMT. As the same relationship was not apparent for the BW waves, this consolidates the view that visionary experiences under DMT correspond to higher amounts of FW waves in particular. Taken together with previous results from visual stimulation experiments independent of DMT ( Pang et al., 2020 ), these data strongly support the principle that cortical travelling waves (and increased FW waves in particular) correlate with the conscious visual experiences, whether induced exogenously (via direct visual stimulation) or endogenously (visionary or hallucinatory experiences).

An external file that holds a picture, illustration, etc.
Object name is elife-59784-fig5.jpg

The first panel shows the correlation between intensity rate and waves amplitude across time-points. Each dot represents a one-minute time-bin from DMT injection, the x-axis reflects the average intensity rating across subjects, and the y-axis indicates the average strength of BW or FW waves across subjects (both correlations p<0.0001). The middle panel shows the correlation coefficients across participants, obtained by correlating the intensity ratings and the waves’ amount separately for each time point. Solid lines show when the amount of waves is significantly larger than zero (always for BW waves, few minutes after DMT injections for FW waves – see Figure 2A ). However, given the limited statistical power (N = 12), and proper correction for multiple testing, correlations did not reach significance at any time point. The last panel shows the correlation coefficients between the visual imagery specific ratings provided at the end of the experiment (i.e. Visual Analogue Scale, see methods) and the net amount of waves (measured when both BW and FW were significantly different than zero, i.e. from minutes 2 to 5): for all 20 items in the questionnaire there was a positive trend between the amount of FW waves and the intensity of visual imagery, as confirmed by a Bayesian t-test against zero (BF for FW waves >> 100). We did not observe this effect in the BW waves (BF = 0.41).

In this study we investigated the effects of the classic serotonergic psychedelic drug DMT on cortical spatio-temporal dynamics typically described as travelling waves ( Muller et al., 2018 ). We analysed EEG signals recorded from 13 participants who kept their eyes closed while receiving drug. Results revealed that, compared with consistent eyes-closed conditions under placebo, eyes-closed DMT is associated with striking changes in cortical dynamics, which are remarkably similar to those observed during actual eyes-open visual stimulation ( Alamia and VanRullen, 2019 ; Pang et al., 2020 ). Specifically, we observed a reduction in BW waves, and increase in FW ones, as well as an overall decrease in α band (8–12 Hz) oscillatory frequencies ( Timmermann et al., 2019 ). Moreover, increases in the amount of FW waves correlated positively with real-time ratings of the subjective intensity of the drug experience as well as post-hoc ratings of visual imagery, suggesting a clear relationship between travelling waves and a distinct and novel type of conscious experience.

Relation to previous findings

Initiated by the discovery of mescaline, and catalysed by the discovery of LSD, Western medicine has explored the scientific value and therapeutic potential of psychedelic compounds for over a century ( Carhart-Harris, 2018 ; Schoen, 1964 ; Strassman, 1995b ). DMT has been evoking particular interest in recent decades, with new studies into its basic pharmacology ( Dean et al., 2019 ), endogenous function ( Barker et al., 2012 ) and effects on cortical activity in rats ( Artigas et al., 2016 ; Riga et al., 2014 ) and humans ( Daumann et al., 2010 ; de Araujo et al., 2012 ; Valle et al., 2016 ). There has been a surprising dearth of resting-state human neuroimaging studies involving pure DMT ( Palhano-Fontes et al., 2015 ; Timmermann et al., 2019 ) which, given its profound and basic effects on conscious awareness, could be viewed as a scientific oversight.

Previous work involving ayahuasca and BOLD fMRI found increased visual cortex BOLD signal under the drug vs placebo while participants engaged in an eyes-closed imagery task – a result that was interpreted as consistent with the ‘visionary’ effects of ayahuasca ( de Araujo et al., 2012 ). Despite some initial debate ( Bartolomeo, 2008 ), it is now generally accepted that occipital cortex becomes activated during visual imagery ( Fulford et al., 2018 ; Pearson, 2019 ). Placing these findings into the context of previous work demonstrating increased FW travelling waves during direct visual perception ( Alamia and VanRullen, 2019 ; Pang et al., 2020 ), our present findings of increased FW waves under DMT correlating with visionary experiences lend significant support to the notion that DMT/ayahuasca – and perhaps other psychedelics – engage the visual apparatus in a fashion that is consistent with actual exogenously driven visual perception. Future work could extend this principle to other apparently endogenous generated visionary experiences such as dream visions and other hallucinatory states. We would hypothesize a consistent favouring of FW waves during these states. If consistent mechanisms were also found to underpin hallucinatory experiences in other sensory modalities – such as the auditory one, a basic principle underlying sensory hallucinations might be established.

Pharmacological considerations

As a classic serotonergic psychedelic drug, DMT’s signature psychological effects are likely mediated by stimulation of the serotonin 2A (5-HT2A) receptor subtype. As with all other classic psychedelics ( Nichols, 2016 ) the 5-HT2A receptor has been found to be essential for the full signature psychological and brain effects of Ayahuasca ( Valle et al., 2016 ). In addition to its role in mediating altered perceptual experiences under psychedelics, the 5-HT2A receptor has also been linked to visual hallucinations in neurological disorders, with a 5-HT2A receptor inverse agonist having been licensed for hallucinations and delusions in Parkinson’s disease with additional evidence for its efficacy in reducing consistent symptoms in Alzheimer’s disease ( Ballard et al., 2018 ). Until recently, a systems level mechanistic account of the role of 5-HT2A receptor agonism in visionary or hallucinatory experiences has, however, been lacking.

Predictive coding and psychedelics

There is a wealth of evidence that Bayesian or predictive mechanisms play a fundamental role in cognitive and perceptual processing ( den Ouden et al., 2012 ; Kok and De Lange, 2015 ) and our understanding of the functional architecture underlying such processing is continually being updated ( Alamia and VanRullen, 2019 ; Friston, 2018 ). According to predictive coding ( Huang and Rao, 2011 ), the brain strives to be a model of its environment. More specifically, based on the assumption that the cortex is a hierarchical system – message passing from higher cortical levels is proposed to encode predictions about the activity of lower levels. This mechanism is interrupted when predictions are contradicted by the lower-level activity (‘prediction error’) – in which case, information passes up the cortical hierarchy where it can update predictions. Predictive coding has recently served as a guiding framework for explaining the psychological and functional brain effects of psychedelic compounds ( Carhart-Harris and Friston, 2019 ; Pink-Hashkes et al., 2017 ). According to one model ( Carhart-Harris and Friston, 2019 ), psychedelics decrease the precision- weighting of top-down priors, thereby liberating bottom-up information flow. Various aspects of the multi-level action of psychedelics are consistent with this model, such as the induction of asynchronous neuronal discharge rates in cortical layer 5 ( Celada et al., 2008 ), reduced alpha oscillations ( Carhart-Harris et al., 2016 ; Muthukumaraswamy et al., 2013 ) increased signal complexity ( Schartner et al., 2017 ; Timmermann et al., 2019 ) and the breakdown of large-scale intrinsic networks ( Carhart-Harris et al., 2016 ).

Recent empirically supported modelling work has lent support to assumptions that top-down predictions and bottom-up prediction-errors are encoded in the direction of propagating cortical travelling waves ( Alamia and VanRullen, 2019 ). Specifically, these simulations demonstrated that a minimal predictive coding model implementing biologically plausible constraints (i.e. temporal delays in the communication between regions and time constants) generates alpha-band travelling waves, which propagate from frontal to occipital regions and vice versa, depending on the ‘cognitive states’ of the model (input-driven vs. prior-driven), as confirmed by EEG data in healthy participants (in that case, processing visual stimuli vs. closed-eyes resting state).

The view that predictive coding could be the underlying principle explaining both the propagation of alpha-band travelling waves and the neural changes induced by psychedelics opened-up a tantalizing opportunity for testing assumptions both about the nature of travelling waves and how they should be modulated by psychedelics ( Carhart-Harris and Friston, 2019 ). Although we are restricted to speculation by the lack of direct experimental manipulation of top-down and bottom-up sensory inputs, our prior assumptions were so emphatically endorsed by the data, including how propagation-shifts related to subjective experience, that, in-line with prior hypotheses and motivations for the analyses, we were persuaded to infer about both the functional relevance of cortical travelling waves and brain action of psychedelics. Additional studies manipulating bottom-up and top-down analysis of sensory inputs with alternative perceptual designs will be required to confirm the relation between predictive coding, alpha-band oscillatory travelling waves and psychedelics states. Moreover, future studies can now be envisioned to examine how these assumptions translate to other phenomena such as non-drug induced visionary and hallucinatory states.

The present analyses were applied to the first EEG data on the effects of DMT on human resting- state brain activity. In-line with a specific prior hypothesis, clear evidence was found of a shift in cortical travelling waves away from the normal basal predominance of backward waves and towards the predominance of forward waves – remarkably similar to what has been observed during eyes-open visual stimulation. Moreover, the increases in forward waves correlated positively with both the general intensity of DMT’s subjective effects, as well as its more specific effects on eyes-closed visual imagery. These findings have specific and broad implications: for the brain mechanisms underlying the DMT/psychedelic state as well as conscious visual perception more fundamentally.

Materials and methods

Participants and experimental procedure.

In this study we analysed a dataset presented in a previous publication ( Timmermann et al., 2019 ), to address a very different scientific question using another analytical approach. Consequently, the information reported in this and the next paragraphs overlaps with the previous study (to which we refer the reader for additional details). Thirteen participants took part in this study (six females, age 34.4 ± 9.1 SD), sample size was chosen based on prior EEG and MEG studies and effect sizes with similar compounds. All participants provided written informed consent, and the study was approved by the National Research Ethics (NRES) Committee London – Brent and the Health Research Authority. The study was conducted in-line with the Declaration of Helsinki and the National Health Service Research Governance Framework.

Participants were carefully screened before joining the experiments. A medical doctor conducted physical examination, electrocardiogram, blood pressure and routine blood tests. A successful psychiatric interview was necessary to join the experiment. Other exclusion criteria were (1) under 18 years of age, (2) having no previous experience with psychedelic drugs, (3) history of diagnosed psychiatric illnesses, (4) excessive use of alcohol (more than 40 units per week). The day before the experiment a urine and pregnancy test (when applicable) were performed, together with a breathalyzer test.

Participants attended two sessions, in the first one, they received placebo, while DMT was administered in the second session. We employed a fixed-order, single blind design considering that psychedelics have been shown to induce lasting psychological changes ( Maclean et al., 2011 ), which could have led to confounding effects on the following session if DMT had been administered in the first session. Additionally, we aimed at ensuring familiarity with the research environment and the study team before providing the psychedelics compound. Given the lack of human data with DMT, progressively increasing doses were provided to different participants (four different doses were used: 7, 14, 18 and 20 mg, to 3, 4, 1, and 5 successive participants, respectively). EEG signals were recorded before and up to 20 min after drug delivery. Participants rested in a semi-supine position with their eyes closed during the duration of the whole experiment. The eyes-closed instruction was confirmed by visual inspection of the participants during dosing. At each minute, participants provided an intensity rating, while blood samples were taken at given time-points (the same for placebo and DMT conditions) via a cannula inserted in the participants’ arm. One day after the DMT session, participants reported their subjective experience completing an interview composed of several questionnaires (see Timmermann et al., 2019  for details). In this study we focused on the Visual Analogue Scale values.

EEG preprocessing

EEG signals were recorded using a 32-channels Brainproduct EEG system sampling at 1000 Hz. A high-pass filter at 0.1 Hz and an anti-aliasing low-pass filter at 450 Hz were applied before applying a band-pass filter at 1–45 Hz. Epochs with artifacts were manually removed upon visual inspection. Independent-component analysis (ICA) was performed and components corresponding to eye-movement and cardiac-related artifacts were removed from the EEG signal. The data were re-referenced to the average of all electrodes. All the preprocessing was carried out using the Fieldtrip toolbox ( Oostenveld et al., 2011 ), while the following analysis were run using custom scripts in MATLAB.

Waves quantification

We epoched the preprocessed EEG signals in 1 s windows, sliding with a step of 500 ms (see Figure 1 ). For each time-window, we then arranged a 2D time-electrode map composed of five electrodes (i.e. Oz, POz, Pz, Cz, FCz). From each map we computed the 2D Fast Fourier Transform (2DFFT – Figure 1 ), from which we extracted the maximum value in the upper and lower quadrants, representing respectively the power of forward (FW) and backward (BW) waves. We also performed the same procedure 100 times after having randomised the electrodes’ order: the surrogate 2D-FFT spectrum has the same temporal frequency content overall, but the spatial information is disrupted, and the information about the wave directionality is lost. In such a manner we obtained the null or surrogate measures, namely FWss and BWss, whose values are the average over the 100 repetitions (see Figure 1 ). Eventually, we computed the actual amount of waves in decibel (dB), considering the log-ratio between the actual and the surrogate values:

It is worth noting that this value represents the amount of significant waves against the null distribution, that is against the hypothesis of having no FW or BW waves.

Statistical analysis

All the analyses regarding the EEG signals were performed in MATLAB. Bayesian analyses were run in JASP ( Team, 2018 ). We ran separate Bayesian ANOVA for FW and BW conditions, and we considered as factors the time of injection (pre-post, see Figure 2A ) and drug condition (DMT vs Placebo). Subjects were considered to account for random factors. Regarding the minute-by-minute analysis ( Figure 2A , right panels), we performed a t-test at each time-bin against zero, and we corrected all the p-values according to the False Discovery Rate ( Benjamini and Yekutieli, 2005 ). All data and code to perform the analysis are available at https://osf.io/wujgp/ .

Acknowledgements

We dedicate this paper to the memory of Jordi Riba , a gracious man and pioneering psychedelic researcher.

Funding Statement

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Contributor Information

Virginie van Wassenhove, CEA, DRF/I2BM, NeuroSpin; INSERM, U992, Cognitive Neuroimaging Unit, France.

Timothy E Behrens, University of Oxford, United Kingdom.

Funding Information

This paper was supported by the following grants:

  • Alexander Mosley Charitable Trust to Robin L Carhart-Harris.
  • Ad Astra Chandaria Foundation to Robin L Carhart-Harris.
  • CRCNS ANR-NSF ANR-19-NEUC-0004 to Rufin VanRullen.
  • ANITI (Artificial and Natural Intelligence Toulouse Institute) Research Chair ANR-19-PI3A-0004 to Rufin VanRullen.
  • Comision Nacional de Investigacion Cientifica y Tecnologica de Chile to Christopher Timmermann.

Additional information

No competing interests declared.

Software, Formal analysis, Visualization, Methodology, Writing - original draft.

Conceptualization, Data curation, Formal analysis, Methodology, Writing - review and editing.

Conceptualization, Data curation, Supervision, Funding acquisition.

Formal analysis, Supervision, Funding acquisition, Methodology, Writing - review and editing.

Conceptualization, Data curation, Supervision, Funding acquisition, Project administration, Writing - review and editing.

Human subjects: All participants provided written informed consent, and the study was approved by the National Research Ethics (NRES) Committee London - Brent and the Health Research Authority (16/LO/0897). The study was conducted in line with the Declaration of Helsinki and the National Health Service Research Governance Framework.

Additional files

Transparent reporting form, data availability.

  • Alamia A, VanRullen R. Alpha oscillations and traveling waves: signatures of predictive coding? PLOS Biology. 2019; 17 :e3000487. doi: 10.1371/journal.pbio.3000487. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Alexander DM, Trengove C, Wright JJ, Boord PR, Gordon E. Measurement of phase gradients in the EEG. Journal of Neuroscience Methods. 2006; 156 :111–128. doi: 10.1016/j.jneumeth.2006.02.016. [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Alexander DM, Jurica P, Trengove C, Nikolaev AR, Gepshtein S, Zvyagintsev M, Mathiak K, Schulze-Bonhage A, Ruescher J, Ball T, van Leeuwen C. Traveling waves and trial averaging: the nature of single-trial and averaged brain responses in large-scale cortical signals. NeuroImage. 2013; 73 :95–112. doi: 10.1016/j.neuroimage.2013.01.016. [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Alexander DM, Ball T, Schulze-Bonhage A, van Leeuwen C. Large-scale cortical travelling waves predict localized future cortical signals. PLOS Computational Biology. 2019; 15 :e1007316. doi: 10.1371/journal.pcbi.1007316. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Artigas F, Riga M, Celada P. The serotonergic hallucinogen 5-MeO-DMT disrupts cortical activity in rodents. European Neuropsychopharmacology. 2016; 26 :S120–S121. doi: 10.1016/S0924-977X(16)30890-2. [ CrossRef ] [ Google Scholar ]
  • Ballard C, Banister C, Khan Z, Cummings J, Demos G, Coate B, Youakim JM, Owen R, Stankovic S, ADP Investigators Evaluation of the safety, tolerability, and efficacy of pimavanserin versus placebo in patients with Alzheimer's disease psychosis: a phase 2, randomised, placebo-controlled, double-blind study. The Lancet Neurology. 2018; 17 :213–222. doi: 10.1016/S1474-4422(18)30039-5. [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Barker SA, McIlhenny EH, Strassman R. A critical review of reports of endogenous psychedelic N, N-dimethyltryptamines in humans: 1955-2010. Drug Testing and Analysis. 2012; 4 :617–635. doi: 10.1002/dta.422. [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Bartolomeo P. The neural correlates of visual mental imagery: an ongoing debate. Cortex. 2008; 44 :107–108. doi: 10.1016/j.cortex.2006.07.001. [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Benjamini Y, Yekutieli D. False discovery Rate–Adjusted Multiple Confidence Intervals for Selected Parameters. Journal of the American Statistical Association. 2005; 100 :71–81. doi: 10.1198/016214504000001907. [ CrossRef ] [ Google Scholar ]
  • Buckholtz NS, Boggan WO. Monoamine oxidase inhibition in brain and liver produced by beta-carbolines: structure-activity relationships and substrate specificity. Biochemical Pharmacology. 1977; 26 :1991–1996. doi: 10.1016/0006-2952(77)90007-7. [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Carhart-Harris RL, Muthukumaraswamy S, Roseman L, Kaelen M, Droog W, Murphy K, Tagliazucchi E, Schenberg EE, Nest T, Orban C, Leech R, Williams LT, Williams TM, Bolstridge M, Sessa B, McGonigle J, Sereno MI, Nichols D, Hellyer PJ, Hobden P, Evans J, Singh KD, Wise RG, Curran HV, Feilding A, Nutt DJ. Neural correlates of the LSD experience revealed by multimodal neuroimaging. PNAS. 2016; 113 :4853–4858. doi: 10.1073/pnas.1518377113. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Carhart-Harris RL. Serotonin, psychedelics and psychiatry. World Psychiatry. 2018; 17 :358–359. doi: 10.1002/wps.20555. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Carhart-Harris RL, Friston KJ. REBUS and the anarchic brain: toward a unified model of the brain action of psychedelics. Pharmacological Reviews. 2019; 71 :316–344. doi: 10.1124/pr.118.017160. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Celada P, Puig MV, Díaz-Mataix L, Artigas F. The hallucinogen DOI reduces low-frequency oscillations in rat prefrontal cortex: reversal by antipsychotic drugs. Biological Psychiatry. 2008; 64 :392–400. doi: 10.1016/j.biopsych.2008.03.013. [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Christian ST, Harrison R, Quayle E, Pagel J, Monti J. The in vitro identification of dimethyltryptamine (DMT) in mammalian brain and its characterization as a possible endogenous neuroregulatory agent. Biochemical Medicine. 1977; 18 :164–183. doi: 10.1016/0006-2944(77)90088-6. [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Daumann J, Wagner D, Heekeren K, Neukirch A, Thiel CM, Gouzoulis-Mayfrank E. Neuronal correlates of visual and auditory alertness in the DMT and ketamine model of psychosis. Journal of Psychopharmacology. 2010; 24 :1515–1524. doi: 10.1177/0269881109103227. [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • de Araujo DB, Ribeiro S, Cecchi GA, Carvalho FM, Sanchez TA, Pinto JP, de Martinis BS, Crippa JA, Hallak JE, Santos AC. Seeing with the eyes shut: neural basis of enhanced imagery following ayahuasca ingestion. Human Brain Mapping. 2012; 33 :2550–2560. doi: 10.1002/hbm.21381. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Dean JG, Liu T, Huff S, Sheler B, Barker SA, Strassman RJ, Wang MM, Borjigin J. Biosynthesis and extracellular concentrations of N,N-dimethyltryptamine (DMT) in mammalian brain. Scientific Reports. 2019; 9 :45812-w. doi: 10.1038/s41598-019-45812-w. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • den Ouden HE, Kok P, de Lange FP. How prediction errors shape perception, attention, and motivation. Frontiers in Psychology. 2012; 3 :1–12. doi: 10.3389/fpsyg.2012.00548. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Freeman WJ, Barrie JM. Analysis of spatial patterns of phase in neocortical gamma EEGs in rabbit. Journal of Neurophysiology. 2000; 84 :1266–1278. doi: 10.1152/jn.2000.84.3.1266. [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Friston K. Does predictive coding have a future? Nature Neuroscience. 2018; 21 :1019–1021. doi: 10.1038/s41593-018-0200-7. [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Friston KJ. Waves of prediction. PLOS Biology. 2019; 17 :e3000426. doi: 10.1371/journal.pbio.3000426. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Fulford J, Milton F, Salas D, Smith A, Simler A, Winlove C, Zeman A. The neural correlates of visual imagery vividness - An fMRI study and literature review. Cortex. 2018; 105 :26–40. doi: 10.1016/j.cortex.2017.09.014. [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Halgren M, Ulbert I, Bastuji H, Fabó D, Erőss L, Rey M, Devinsky O, Doyle WK, Mak-McCully R, Halgren E, Wittner L, Chauvel P, Heit G, Eskandar E, Mandell A, Cash SS. The generation and propagation of the human alpha rhythm. PNAS. 2019; 116 :23772–23782. doi: 10.1073/pnas.1913092116. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Huang Y, Rao RPN. Predictive coding. Wiley Interdisciplinary Reviews: Cognitive Science. 2011; 2 :580–593. doi: 10.1002/wcs.142. [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Kok P, De Lange FP. Predictive coding in sensory cortex. In: Forstmann B, Wagenmakers E. J, editors. An Introduction to Model-Based Cognitive Neuroscience. Springer; 2015. pp. 221–244. [ CrossRef ] [ Google Scholar ]
  • Lozano-Soldevilla D, VanRullen R. The hidden spatial dimension of alpha: 10-hz perceptual echoes propagate as periodic traveling waves in the human brain. Cell Reports. 2019; 26 :374–380. doi: 10.1016/j.celrep.2018.12.058. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Maclean KA, Johnson MW, Griffiths RR. Psilocybin lead to increases in the personality domain of openness. J Psychoph. 2011; 25 :1453–1461. doi: 10.1177/0269881111420188. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Muller L, Reynaud A, Chavane F, Destexhe A. The stimulus-evoked population response in visual cortex of awake monkey is a propagating wave. Nature Communications. 2014; 5 :4675. doi: 10.1038/ncomms4675. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Muller L, Chavane F, Reynolds J, Sejnowski TJ. Cortical travelling waves: mechanisms and computational principles. Nature Reviews Neuroscience. 2018; 19 :255–268. doi: 10.1038/nrn.2018.20. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Muthukumaraswamy SD, Carhart-Harris RL, Moran RJ, Brookes MJ, Williams TM, Errtizoe D, Sessa B, Papadopoulos A, Bolstridge M, Singh KD, Feilding A, Friston KJ, Nutt DJ. Broadband cortical desynchronization underlies the human psychedelic state. Journal of Neuroscience. 2013; 33 :15171–15183. doi: 10.1523/JNEUROSCI.2063-13.2013. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Nichols DE. Psychedelics. Pharmacological Reviews. 2016; 68 :264–355. doi: 10.1124/pr.115.011478. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Nunez PL. The brain wave equation: a model for the EEG. Mathematical Biosciences. 1974; 21 :279–297. doi: 10.1016/0025-5564(74)90020-0. [ CrossRef ] [ Google Scholar ]
  • Nunez PL. Toward a quantitative description of large-scale neocortical dynamic function and EEG. Behavioral and Brain Sciences. 2000; 23 :371–398. doi: 10.1017/S0140525X00003253. [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Nunez PL, Srinivasan R. Electric Fields of the Brain: The Neurophysics of EEG. Second Edition. Oxford University Press; 2009. [ CrossRef ] [ Google Scholar ]
  • Nunez PL, Srinivasan R. Neocortical dynamics due to axon propagation delays in cortico-cortical fibers: eeg traveling and standing waves with implications for top-down influences on local networks and white matter disease. Brain Research. 2014; 1542 :138–166. doi: 10.1016/j.brainres.2013.10.036. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Oostenveld R, Fries P, Maris E, Schoffelen JM. FieldTrip: open source software for advanced analysis of MEG, EEG, and invasive electrophysiological data. Computational Intelligence and Neuroscience. 2011; 2011 :1–9. doi: 10.1155/2011/156869. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Palhano-Fontes F, Andrade KC, Tofoli LF, Santos AC, Crippa JA, Hallak JE, Ribeiro S, de Araujo DB. The psychedelic state induced by Ayahuasca modulates the activity and connectivity of the default mode network. PLOS ONE. 2015; 10 :e0118143. doi: 10.1371/journal.pone.0118143. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Pang Z, Alamia A, Vanrullen R. Turning the stimulus on and off dynamically changes the direction of alpha travelling waves. bioRxiv. 2020 doi: 10.1101/2020.04.15.041756. [ CrossRef ]
  • Pearson J. The human imagination: the cognitive neuroscience of visual mental imagery. Nature Reviews Neuroscience. 2019; 20 :624–634. doi: 10.1038/s41583-019-0202-9. [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Pink-Hashkes S, Rooij van, Kwisthout J. Perception is in the details: a predictive coding account of the psychedelic phenomenon. CogSci Annual Conference of the Cognitive Science Society; 2017. pp. 2907–2912. [ Google Scholar ]
  • Riba J, Anderer P, Morte A, Urbano G, Jané F, Saletu B, Barbanoj MJ. Topographic pharmaco-EEG mapping of the effects of the south american psychoactive beverage ayahuasca in healthy volunteers. British Journal of Clinical Pharmacology. 2002; 53 :613–628. doi: 10.1046/j.1365-2125.2002.01609.x. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Riga MS, Soria G, Tudela R, Artigas F, Celada P. The natural hallucinogen 5-MeO-DMT, component of Ayahuasca, disrupts cortical function in rats: reversal by antipsychotic drugs. The International Journal of Neuropsychopharmacology. 2014; 17 :1269–1282. doi: 10.1017/S1461145714000261. [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Sato TK, Nauhaus I, Carandini M. Traveling waves in visual cortex. Neuron. 2012; 75 :218–229. doi: 10.1016/j.neuron.2012.06.029. [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Schartner MM, Carhart-Harris RL, Barrett AB, Seth AK, Muthukumaraswamy SD. Increased spontaneous MEG signal diversity for psychoactive doses of ketamine, LSD and psilocybin. Scientific Reports. 2017; 7 :46421. doi: 10.1038/srep46421. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Schenberg EE, Alexandre JF, Filev R, Cravo AM, Sato JR, Muthukumaraswamy SD, Yonamine M, Waguespack M, Lomnicka I, Barker SA, da Silveira DX. Acute biphasic effects of Ayahuasca. PLOS ONE. 2015; 10 :e0137202. doi: 10.1371/journal.pone.0137202. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Schoen SM. Hallucinogenic drugs and their psychotherapeutic use. American Journal of Psychotherapy. 1964; 18 :338–340. doi: 10.1176/appi.psychotherapy.1964.18.2.338. [ CrossRef ] [ Google Scholar ]
  • Smythies JR, Morin RD, Brown GB. Identification of dimethyltryptamine and O-methylbufotenin in human cerebrospinal fluid by combined gas chromatography/mass spectrometry. Biological Psychiatry. 1979; 14 :549–556. [ PubMed ] [ Google Scholar ]
  • Strassman RJ. Dose-Response study of N,N-Dimethyltryptamine in humans. Archives of General Psychiatry. 1994; 51 :98. doi: 10.1001/archpsyc.1994.03950020022002. [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Strassman RJ. Human psychopharmacology of N,N-dimethyltryptamine. Behavioural Brain Research. 1995a; 73 :121–124. doi: 10.1016/0166-4328(96)00081-2. [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Strassman RJ. Hallucinogenic drugs in psychiatric research and treatment perspectives and prospects. The Journal of Nervous and Mental Disease. 1995b; 183 :127–138. doi: 10.1097/00005053-199503000-00002. [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Strassman R. DMT: The Spirit Molecule. Park Street Press; 2001. [ Google Scholar ]
  • Team J. Computer software; 2018. https://github.com/jasp-stats [ Google Scholar ]
  • Timmermann C, Roseman L, Schartner M, Milliere R, Williams LTJ, Erritzoe D, Muthukumaraswamy S, Ashton M, Bendrioua A, Kaur O, Turton S, Nour MM, Day CM, Leech R, Nutt DJ, Carhart-Harris RL. Neural correlates of the DMT experience assessed with multivariate EEG. Scientific Reports. 2019; 9 :4. doi: 10.1038/s41598-019-51974-4. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • Valle M, Maqueda AE, Rabella M, Rodríguez-Pujadas A, Antonijoan RM, Romero S, Alonso JF, Mañanas MÀ, Barker S, Friedlander P, Feilding A, Riba J. Inhibition of alpha oscillations through serotonin-2A receptor activation underlies the visual effects of ayahuasca in humans. European Neuropsychopharmacology. 2016; 26 :1161–1175. doi: 10.1016/j.euroneuro.2016.03.012. [ PubMed ] [ CrossRef ] [ Google Scholar ]
  • eLife. 2020; 9: e59784.

Decision letter

David murray alexander.

KU Leuven, Belgium

In the interests of transparency, eLife publishes the most substantive revision requests and the accompanying author responses.

Acceptance summary:

In this study, Alamia and colleagues describe the effect of N,N, Dimethyltryptamine DMT on the resting-state dynamics of α traveling waves. DMT is a serotonergic psychedelic drug that elicits vivid hallucinations. The authors tested whether DMT provoke a relative change in strength of forward- and backward- α traveling waves recorded with non-invasive electroencephalography. Specifically, the hypothesis was that traveling waves under DMT may show features of visual perception in the absence of photic stimulation. Indeed, following DMT consumption and during eyes-closed resting state, the authors report an increase of forward-traveling waves, an α power increase (comparable to photic stimulation) and of low-frequency components in the low-range spectrum. The implication of traveling waves are discussed in relation to predictive coding.

Decision letter after peer review:

Thank you for submitting your article "DMT alters cortical travelling waves" for consideration by eLife . Your article has been reviewed by three peer reviewers, one of whom is a member of our Board of Reviewing Editors, and the evaluation has been overseen by Timothy Behrens as the Senior Editor. The following individual involved in review of your submission has agreed to reveal their identity: David Murray Alexander (Reviewer #2).

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

We would like to draw your attention to changes in our revision policy that we have made in response to COVID-19 (https://elifesciences.org/articles/57162). Specifically, when editors judge that a submitted work as a whole belongs in eLife but that some conclusions require a modest amount of additional new data, as they do with your paper, we are asking that the manuscript be revised to either limit claims to those supported by data in hand, or to explicitly state that the relevant conclusions require additional supporting data.

Our expectation is that the authors will eventually carry out the additional experiments and report on how they affect the relevant conclusions either in a preprint on bioRxiv or medRxiv, or if appropriate, as a Research Advance in eLife , either of which would be linked to the original paper.

Alamia and colleagues describe the effect of N,N, Dimethyltryptamine DMT on human resting-state dynamics recorded with non-invasive EEG. DMT is a serotonergic psychedelic drug that elicits vivid hallucinations. The authors propose that DMT provokes a relative change in strength of forward- and backward- α [~10 Hz] traveling waves, indicative of bottom-up and top-down propagations of information. In three main analyses of EEG recordings following the administration of DMT, the authors report an increase of forward-traveling waves, α power (comparable to photic stimulation) and low-frequency components in the low-range spectrum.

Revisions for this paper:

I highlight three main issues that need to be addressed in a revised paper.

1) All three reviewers highlight specific points re. the limitations of the current analysis (e.g. a prior choice of electrodes) and the choice for quantifications of the traveling waves. The authors should clarify their rationale underlying their analytical choices. As suggested by reviewer 3, a section dedicated to "Quantifying travelling waves" may be helpful.

2) All three reviewers raised substantial concerns about the interpretability of scalp level recordings in relation to the generators of the signals as well as the inference that can be made on the traveling pattern. Reviewer 1 raised concerns about the full spectral changes that can be seen and reviewer 3 suggested the possibility of interference patterns.
3) Reviewers 1 and 2 consider the predictive coding hypothesis a far-stretched inference of current results, and thus needs to be refined.

Reviewer #1:

In this study, Alamia and colleagues describe the effect of N,N, Dimethyltryptamine DMT on the resting-state dynamics of α traveling waves. DMT is a serotonergic psychedelic drug that elicits vivid hallucinations. The authors herein propose that DMT provokes a relative change in strength of forward- and backward- α traveling waves recorded with non-invasive EEG. The authors hypothesized that traveling waves under DMT may show features of visual perception, namely an increase of forward going (occipital to frontal) traveling waves during eyes-closed resting-state thus independently of photic inputs. With three main analyses, the authors observe that following DMT, forward-traveling waves increase, α power increase (comparable to photic stimulation) and an overall increase of low-frequency components in the low-range spectrum.

I have several major concerns regarding the reliability of the analysis and subsequent interpretations. One is that while the authors quantified traveling waves, they also report a radical change in the overall spectral fingerprinting of the EEG (Figure 3) and clearly showing that α is largely suppressed following DMT thus largely diminishing the reliability and pertinence of focusing on α traveling wave while low-frequency are largely boosted. Second, the authors report consistent inverse relationships between FW and BW waves, which may simply result from moving dipoles that generate the signals; in light of this, the inverse relation in Figure 5 between FW/BW is not surprising. Finally, would a simpler measure of decrease in α power and increase of low spectral power reveal similar correlations to behavior?

Reviewer #2:

The authors present a sensor level analysis of traveling waves in the EEG, during dosage with DMT or saline. The work is a strong contribution to the study of cortical traveling waves (TWs) due to the pharmacological manipulation, which helps our understanding of the causal role of TWs. This contribution is bolstered by being able to draw parallels with the effects of visual stimulation (or not) during rest, reported in a contemporaneous manuscript. The manuscript will be of general interest to readers of eLife . The manuscript is crisply written.

The conclusions follow from the analysis.

1) The quantitative methods to analyse TWs are rather week, being focussed on direction of travel along the anterior-posterior axis. While these methods are sufficient to support the main conclusions, more could be teased from the experiment by following recent developments in TW quantification.

For example, using peak tracing along the lines of Massimini et al., 2004, would enable the detailed paths of the waves to be traced over the whole recording array, and velocity to be estimated.

Methods exist to estimate the spatial frequency of the waves on the scalp, as well as the proportion of traveling vs. standing waves (Alexander et al., 2016). Likewise, other directional components of the wave trajectory can be assessed by using PCA to create a spatial basis for the waves (Alexander et al., 2006, 2009, 2013).

It seems possible that important features of the data have been missed by limiting the analysis to electrodes FCz to Oz. For example, what if DMT influence and visual stimulation share a common primary direction, as is found, but DMT waves are more left posterior to right anterior and visual stimulation is more right posterior to left anterior (or vice versa)?
2) The sections on predictive coding are only tenuously supported by the data. In particular, I can see no discussion on how directional differences in the α band may be significant in this regard. What about situations where anterior-posterior differences are found in the δ band (Alexander et al., 2006; 2009)? Or if directional results were in another band? Because of the lack of specificity to this discussion, and the lack of explicit tests of this theoretical framework, I suggest these concepts be given a more appropriate weighting (less).
3) An obvious objection to the analysis is that it is sensor level. The authors need to address their reasons for doing this e.g. that source projections destroy real long-range correlations as well as blurring by the scalp and other tissues. See Nunez, 1974; Nunez et al., 1997; Freeman et al., 2000; 2003 and Alexander et al., 2019, for a summary of these issues.

Reviewer #3:

This study of the effects of the drug DMT on the direction and occurrence of EEG traveling waves seems generally plausible to me, although some important aspects are not discussed. I don't have major criticisms. However, as one who has studied EEG traveling and standing waves for many years, I worry that those unfamiliar with EEG wave phenomena may misinterpret some of these results given their partial dependence on the specific experimental methods employed. While I have not read previous papers by these authors that may fill in some of the important gaps, I list below some ideas that any reader interested in EEG waves and their neuro-scientific interpretation must be aware of. A summary paragraph in the section "Quantifying travelling waves" is recommended concerning the following basic concepts that must be understood if the results are to be interpreted correctly.

1) In all but the simplest systems, traveling waves occur in groups (packets) over some range of spatial wavelengths (multiple spatial frequencies, k). This is to be expected in brains, based on both theory and experiment (see Nunez and Srinivasan, 2006; 2014; Nunez, 2000).

2) Any experimental electrode array will be sensitive to only parts of these wave packets, e.g. only waves shorter than the spatial extent of the array and waves longer than twice the electrode separation distance (Nyquist criterion in space) can be resolved. In scalp recordings, the shorter waves may be mostly removed by volume conduction.

3) As a consequence of #2, waves recorded directly from the cortex (as indicated in several recent studies) will emphasize shorter waves than the scalp recorded waves. In the case of small cortical arrays, the ECoG overlap with scalp data may be minimal. Thus, the different estimated wave properties (including propagation direction) need not agree.

4) When waves are traveling in multiple directions at nearly the same time in "closed" systems (e.g., the cortical/white matter), there are only two possible results. Either the waves must damp out or they combine (interfere) to form standing waves (e.g. α waves traveling both forward and backward). One expects that the actual behavior depends on brain state, including the influence of drugs (see Nunez and Srinivasan, 2006; 2014; Nunez, 2000).

Author response

Revisions for this paper: I highlight three main issues that need to be addressed in a revised paper. 1) All three reviewers highlight specific points re. the limitations of the current analysis (e.g. a prior choice of electrodes) and the choice for quantifications of the traveling waves. The authors should clarify their rationale underlying their analytical choices. As suggested by reviewer 3, a section dedicated to "Quantifying travelling waves" may be helpful.

We have now acknowledged the limitations of the current analyses, and we performed several additional steps to improve on our previous approach. As explained in detail in what follows (and in the revised manuscript) we included a new analysis considering different lines of electrodes, spanning from posterior to anterior, but also left to right brain regions. We also performed an additional complementary analysis based on the suggestions of reviewer 2, showing similar results as our original analysis. Finally, as suggested by reviewer 3, we integrated the current “Quantifying travelling waves” paragraph with his generous suggestions, improving the readability of the manuscript to a non-specialized audience. We believe that these modifications will satisfy both the editor and all the reviewers.

We agree with the concerns raised by both reviewers, and we have included an additional analysis (now Figure 3B in the revised manuscript) that addresses specifically these concerns. Regarding the spectral changes, our new analysis avoids choosing a priori a single frequency band (i.e. the one corresponding to the global maximum in the 2D-FFT) but instead analyze the changes in all the spectrum. This novel approach, besides confirming our previous results, provides a fuller view of the overall changes in the spectral pattern induced by DMT. Concerning the source generators of the travelling waves pattern, we discuss this point in the revised manuscript, arguing that a sensor analysis is more appropriate in this case because it circumvents some limitations related specifically to source analysis (e.g. source projections impair long-range connections). Finally, as shown in Figure 4, backward and forward waves were negatively correlated on a trial by trial basis, which will tend to limit the possibility suggested by reviewer 3 of having interference patterns (resulting in standing waves).

We addressed carefully this point by giving overall less weight to the Predictive Coding hypothesis in the Discussion, as suggested by both reviewers. Additionally, as suggested specifically by reviewer 2, we clarify in the revised Discussion the link between Predictive Coding, α oscillations and travelling waves, and the motivation behind our original hypothesis and the relationship with the present results. More precisely, we previously demonstrated how a model based on Predictive Coding principles and implementing biologically plausible constraints gives rise to α-band travelling waves, whose direction of propagation depends on the “cognitive” state of the model/subject (FW during visual stimulation, BW during closed-eyes resting state). Then, starting from the premise that psychedelics disrupt prior distributions encoded in hierarchically high-level properties of brain function (Carhart-Harris and Friston, 2019), we formulated the specific hypothesis that DMT could specifically disrupt α-band travelling waves, enhancing their feed-forward propagation while decreasing the feed-back direction. All in all, we found it remarkable that such a specific hypothesis received such clear support in the data. However, we understand that interpretations can always be queried and more work is needed to scrutinize the one we offer based on our specific prior hypothesis. We have now rephrased our interpretation substantially in the revised version of the manuscript, to soften our conclusions and emphasize the need for more research.

Reviewer #1: In this study, Alamia and colleagues describe the effect of N,N, Dimethyltryptamine DMT on the resting-state dynamics of α traveling waves. DMT is a serotonergic psychedelic drug that elicits vivid hallucinations. The authors herein propose that DMT provokes a relative change in strength of forward- and backward- α traveling waves recorded with non-invasive EEG. The authors hypothesized that traveling waves under DMT may show features of visual perception, namely an increase of forward going (occipital to frontal) traveling waves during eyes-closed resting-state thus independently of photic inputs. With three main analyses, the authors observe that following DMT, forward-traveling waves increase, α power increase (comparable to photic stimulation) and an overall increase of low-frequency components in the low-range spectrum. I have several major concerns regarding the reliability of the analysis and subsequent interpretations. One is that while the authors quantified traveling waves, they also report a radical change in the overall spectral fingerprinting of the EEG (Figure 3) and clearly showing that α is largely suppressed following DMT thus largely diminishing the reliability and pertinence of focusing on α traveling wave while low-frequency are largely boosted. Second, the authors report consistent inverse relationships between FW and BW waves, which may simply result from moving dipoles that generate the signals; in light of this, the inverse relation in Figure 5 between FW/BW is not surprising. Finally, would a simpler measure of decrease in α power and increase of low spectral power reveal similar correlations to behavior?

We thank the reviewer for raising these important considerations. Regarding the relationship between the spectral changes in the EEG and the changes in the amount of waves, we performed an additional analysis quantifying these changes as a function of each frequency (i.e. extracting in the 2D-FFT the maximum value separately for each frequency). The first row of Author response image 1 (integrated in the revised version of the manuscript as figure 3B) shows the amount of FW and BW waves (in dB –as compared to the surrogate distribution) before and after DMT or Placebo infusion (i.e. Pre and Post). The second row shows the difference in the amount of waves between DMT and Placebo for each frequency. Interestingly the largest changes occur in the α band frequency for both forward and backward waves, even though after correcting for multiple comparison we found a significant reduction only in the BW α band. This analysis shows that, as suggested by the reviewer, the changes in the spectral fingerprint of the EEG do influence the waves’ propagation in several frequencies, but the largest changes systematically occur in the α band. This additional analysis has been introduced in the Results section (paragraph: “Does DMT influence the frequency of travelling waves?”)

An external file that holds a picture, illustration, etc.
Object name is elife-59784-resp-fig1.jpg

The lower panels show the difference between DMT- and placebo-induced waves for each condition. As shown in the previous analysis, DMT induces an overall decrease of the waves’ amplitude, especially pronounced (and significant) in the α band BW waves, with the notable exception of FW waves in the α range, where an DMT-induced increase is observed.

Regarding the relationship between FW and BW waves, our analysis –shown in Figure 4 of the manuscript- suggests that after DMT –or during photic visual stimulation- FW and BW waves do not co-occur simultaneously, but tend to alternate temporally, as revealed by the negative correlation. We agree with the reviewer that moving dipoles could be responsible for the generation of these signals, as reported in an in-depth analysis of a previous paper investigating the source localization of similar waves patterns (Lozano-Soldevilla and VanRullen, 2019). In line with a similar comment from reviewer 2, we added in the manuscript a reference to source vs sensors analysis (“In addition, it is important to note that our waves’ analysis focuses on the sensor level, as source projections present a number of important limitations, such as impairing long-range connections, as well as smearing of signals due to scalp interference (Alexander et al., 2019; Freeman and Barrie, 2000; Nunez, 1974)”). Lastly, a previous analysis on the same dataset (Timmermann et al., 2019) identified a correlation between theta- and δ-band spectral power changes and subjective behavior, but not for changes in the α range. This suggests that the correlation reported in Figure 5 is not a direct consequence of changes in the EEG spectral power.

Reviewer #2: The authors present a sensor level analysis of traveling waves in the EEG, during dosage with DMT or saline. The work is a strong contribution to the study of cortical traveling waves (TWs) due to the pharmacological manipulation, which helps our understanding of the causal role of TWs. This contribution is bolstered by being able to draw parallels with the effects of visual stimulation (or not) during rest, reported in a contemporaneous manuscript. The manuscript will be of general interest to readers of eLife. The manuscript is crisply written. The conclusions follow from the analysis. Concerns: 1) The quantitative methods to analyse TWs are rather week, being focussed on direction of travel along the anterior-posterior axis. While these methods are sufficient to support the main conclusions, more could be teased from the experiment by following recent developments in TW quantification. For example, using peak tracing along the lines of Massimini et al., 2004, would enable the detailed paths of the waves to be traced over the whole recording array, and velocity to be estimated. Methods exist to estimate the spatial frequency of the waves on the scalp, as well as the proportion of traveling vs. standing waves (Alexander et al., 2016). Likewise, other directional components of the wave trajectory can be assessed by using PCA to create a spatial basis for the waves (Alexander et al., 2006, 2009, 2013).

We thank the reviewer for his useful suggestions. As correctly noticed, our current method to detect travelling waves focuses exclusively on the Anterior-Posterior axis, in line with our previous studies (Alamia and VanRullen, 2019; Lozano-Soldevilla and VanRullen, 2019; Pang et al., 2020). However, we agree that more can be inferred from the data from other electrodes (see next point and comment to reviewer 1 for waves’ quantification on other axes). We applied a method similar to Alexander et al., 2006, 2009 and 2016, thus considering the phase of the signal over the entire array of electrodes. Specifically, we computed the phase of the α band-pass filtered signals pre- and post- DMT infusion, and referenced it to the central electrode Cz. The relative phase thus describes the propagation of the wave as compared to this electrodes: positive lags (in yellow in the Author response image 2 ) characterize earlier components, whereas negative lags (in blue) are associated with signals lagging behind. Author response image 2 summarizes the results in all conditions.

An external file that holds a picture, illustration, etc.
Object name is elife-59784-resp-fig2.jpg

Reassuringly, the pattern of results confirms the disruption of the top-down flow, counterbalanced by a bottom-up component, specifically after the infusion of DMT, in line with our original analysis.

Interestingly, we observed that after placebo, the typical top-down propagation of α-band waves remains unaltered, whereas after DMT, waves propagate both FW and BW, as revealed by an overall phase distribution around zero. Overall these results confirmed the one obtained with the 2D-FFT approach. We opted for keeping the latter for consistency with our previous studies (but we mentioned this result in the revised manuscript along with the references)

“Besides, in line with previous work on travelling waves (Alexander et al., 2013, 2006), an additional analysis based on relative phases of the α band-pass signals over all channels confirmed the same results, with DMT disrupting the typical top-down propagation of α-band waves (not shown).”

We agree with the reviewer that the approach used by (Massimini et al., 2004), would allow to identify the detailed path of the waves, and potentially their velocity. However, in their work, Massimini and colleagues targeted slow 1Hz waves (the signal was actually low-pass filtered at 4Hz); for each cycle waves were tracked based on the localization of the main (negative) peak whose voltage was below a threshold of -80 V. This approach, which provides reliable results for low-frequency waves, may present some non-trivial additional limitations when applied to higher frequencies. Specifically, the identification of each peak/cycle may not be straightforward for higher frequencies (e.g. α band); in Massimini et al. the time window used to identify the peak spanned between +/-800ms to the earliest peak, but such a window should be proportionally much shorter to isolate single peaks in the α range, and thus be increasingly more sensitive to jittered noise. We therefore favored the 2D-FFT approach, which –despite its own limitations- seemed more suitable to describe waves with higher temporal frequencies. Finally, regarding the waves speed, it is possible to estimate their velocity from the 2D-FFT, considering both spatial and temporal frequencies as shown in our previous study (Alamia and VanRullen, 2019). The reported results are consistent with the speed recorded for cortical waves (macroscopic scale, speed ~1.5 – 2.0 m/s (Muller et al., 2018)).

We agree with the reviewer that the choice of the midline electrodes supports the main conclusion but prevents a broader view on the waves’ dynamic at the cortical level. Accordingly, and in line with the concerns of reviewer 1, we explored different lines of electrodes, to identify other axes of propagation. As shown in Figure 2—figure supplement 2, comparing PRE and POST DMT infusion reveals an increase of waves propagating from posterior to anterior regions considering an array of electrodes arranged from right posterior to left anterior (diag1 in Figure 2—figure supplement 2, Bayesian t-test BF=4.059, error=0.002%) and from left posterior to right anterior (diag2 in Figure 2—figure supplement 2, Bayesian t-test BF=4.848, error=0.0001%), similarly to the results obtained considering the main posterior-anterior axis (Bayesian t-test BF=5.4, error=0.001%). Additionally, we revealed a significant amount of waves (larger than 0 dB) propagating along the coronal line of electrodes (i.e. leftward and rightward), but those waves were not influenced by DMT infusion (for both leftward and rightward waves BF<0.4, error~0.02%). We included these analyses in the Results section and as Figure 2—figure supplement 2.

“In order to explore different propagation axes than the midline, we ran the same analysis on one array of electrodes running from posterior right to anterior left regions, and one from posterior left to anterior right ones: in both cases we obtained similar results as for the midline electrodes, that is, an increase and a decrease of FW and BW waves respectively following DMT infusion (see Figure 2—figure supplement 2).This suggests that the waves’ propagation spread to most posterior and frontal recording channels As a control, we additionally demonstrated that waves propagating from leftward to rightward regions (and vice versa) were not affected by DMT, as predicted by our hypothesis (see Figure 2—figure supplement 2).”

We agree and thank the reviewer for pointing out this shortcoming in the Discussion. The focus on the α-band originates from our previous study (Alamia and VanRullen, 2019) in which we demonstrated how a minimal Predictive Coding model implementing biologically plausible constraints (i.e. temporal delays in the communication between regions and time constants) generates α-band travelling waves whose direction of propagation is matched by experimental data. This result was the starting hypothesis that motivated the investigation of α-band travelling waves after DMT-infusion, under the hypothesis of the REBUS model (psychedelics disrupt prior distributions in higher brain regions- Carhart-Harris and Friston, 2019). In the revised version of the manuscript, we clarify the link between Predictive Coding, α oscillations and travelling waves. However, we agree with this reviewer (and reviewer 1 and the editor) that the Predictive Coding interpretation may not be directly but only indirectly supported by the current data, and so we have rephrased the relevant section and substantially softened it in the revised version of the manuscript, in accordance with this valid point.

“Specifically, these simulations demonstrated that a minimal Predictive Coding model implementing biologically plausible constraints (i.e. temporal delays in the communication between regions and time constants) generates α-band travelling waves, which propagate from frontal to occipital regions and vice versa, depending on the “cognitive states” of the model (input-driven vs prior-driven), as confirmed by EEG data in healthy participants (in that case, processing visual stimuli vs. closed-eyes resting state). The view that Predictive Coding could be the underlying principle explaining both the propagation of α-band travelling waves, and the neural changes induced by psychedelics opened-up a tantalizing opportunity for testing assumptions both about the nature of travelling waves and how they should be modulated by psychedelics (Carhart-Harris and Friston, 2019). Although we are restricted to speculation by the lack of direct experimental manipulation of top-down and bottom-up sensory inputs, our prior assumptions were so emphatically endorsed by the data, including how propagation-shifts related to subjective experience, that, in-line with prior hypotheses and motivations for the analyses, we were persuaded to infer about both the functional relevance of cortical travelling waves and brain action of psychedelics. Additional studies manipulating bottom-up and top-down analysis of sensory inputs with alternative perceptual designs will be required in order to confirm the relation between Predictive Coding, α-band oscillatory travelling waves and psychedelics states. Moreover, future studies can now be envisioned to examine how these assumptions translate to other phenomena such as non-drug induced visionary and hallucinatory states.”

We thank the reviewer for the useful reminder. We included in the “Quantifying the waves” paragraph of the revised manuscript a few sentence addressing this issue:

“In addition, it’s important to note that our waves’ analysis focuses at the sensor level, as source projections presents few important limitations such as impairing long-range connections, as well as smearing the signals due to the scalp inference (Alexander et al., 2019; Freeman and Barrie, 2000; Nunez, 1974)”

Reviewer #3: This study of the effects of the drug DMT on the direction and occurrence of EEG traveling waves seems generally plausible to me, although some important aspects are not discussed. I don't have major criticisms. However, as one who has studied EEG traveling and standing waves for many years, I worry that those unfamiliar with EEG wave phenomena may misinterpret some of these results given their partial dependence on the specific experimental methods employed. While I have not read previous papers by these authors that may fill in some of the important gaps, I list below some ideas that any reader interested in EEG waves and their neuro-scientific interpretation must be aware of. A summary paragraph in the section "Quantifying travelling waves" is recommended concerning the following basic concepts that must be understood if the results are to be interpreted correctly. 1) In all but the simplest systems, traveling waves occur in groups (packets) over some range of spatial wavelengths (multiple spatial frequencies, k). This is to be expected in brains, based on both theory and experiment (see Nunez and Srinivasan, 2006; 2014; Nunez, 2000). 2) Any experimental electrode array will be sensitive to only parts of these wave packets, e.g. only waves shorter than the spatial extent of the array and waves longer than twice the electrode separation distance (Nyquist criterion in space) can be resolved. In scalp recordings, the shorter waves may be mostly removed by volume conduction. 3) As a consequence of #2, waves recorded directly from the cortex (as indicated in several recent studies) will emphasize shorter waves than the scalp recorded waves. In the case of small cortical arrays, the ECoG overlap with scalp data may be minimal. Thus, the different estimated wave properties (including propagation direction) need not agree. 4) When waves are traveling in multiple directions at nearly the same time in "closed" systems (e.g., the cortical/white matter), there are only two possible results. Either the waves must damp out or they combine (interfere) to form standing waves (e.g. α waves traveling both forward and backward). One expects that the actual behavior depends on brain state, including the influence of drugs (see Nunez and Srinivasan, 2006; 2014; Nunez, 2000).

We are very grateful to the reviewer for her/his overall positive opinion on our work, and her/his useful suggestions. As recommended, we integrated in the “Quantifying travelling waves” paragraph all the points listed above, with the corresponding references. We believe such changes improved the readability of the paper for those unfamiliar with EEG analysis, and hopefully will be satisfying and adequate for the reviewer.

“As demonstrated by both theoretical and experimental evidence (Nunez, 2000; Nunez and Srinivasan, 2014, 2009), in most systems, including the human brain, traveling waves occur in groups (or packets) over some range of spatial wavelengths having multiple spatial and temporal frequencies. Given any configurations of electrodes, only parts of these packets can be successfully detected, i.e. waves shorter than the spatial extent of the array, and waves longer than twice the electrode separation distance (Nyquist criterion in space). In scalp recordings, the shorter waves may be mostly removed by volume conduction. As a consequence, waves recorded directly from the cortex emphasize shorter waves than the scalp recorded waves. Specifically, in the case of small cortical arrays, the overlap between cortical and scalp data may be minimal, and the estimated wave properties (including propagation direction) may differ. Additionally, it is important to consider that when waves are traveling in multiple directions at nearly the same time in "closed" systems (e.g., the cortical/white matter), waves either damp out or interfere with each other to form standing waves (e.g. α waves traveling both forward and backward). It is reasonable to assume that the behavior of these properties will relate to global brain and mind states, and be sensitive to state-altering psychoactive drugs (Nunez, 2000; Nunez and Srinivasan, 2014, 2009).”

Massimini M, Huber R, Ferrarelli F, Hill S, Tononi G. 2004. The sleep slow oscillation as a traveling wave. J Neurosci 24 :6862–6870. doi:10.1523/JNEUROSCI.1318-04.2004

  • Search Menu
  • CNS Injury and Stroke
  • Epilepsy and Sleep
  • Movement Disorders
  • Multiple Sclerosis/Neuroinflammation
  • Neuro-oncology
  • Neurodegeneration - Cellular & Molecular
  • Neuromuscular Disease
  • Neuropsychiatry
  • Pain and Headache
  • Advance articles
  • Editor's Choice
  • Author Guidelines
  • Submission Site
  • Why publish with this journal?
  • Open Access
  • About Brain
  • Editorial Board
  • Advertising and Corporate Services
  • Journals Career Network
  • Self-Archiving Policy
  • Dispatch Dates
  • Terms and Conditions
  • Journals on Oxford Academic
  • Books on Oxford Academic

Issue Cover

Article Contents

Introduction, materials and methods, acknowledgements, competing interests, supplementary material.

  • < Previous

Interictal discharges in the human brain are travelling waves arising from an epileptogenic source

ORCID logo

  • Article contents
  • Figures & tables
  • Supplementary Data

Joshua M Diamond, C Price Withers, Julio I Chapeton, Shareena Rahman, Sara K Inati, Kareem A Zaghloul, Interictal discharges in the human brain are travelling waves arising from an epileptogenic source, Brain , Volume 146, Issue 5, May 2023, Pages 1903–1915, https://doi.org/10.1093/brain/awad015

  • Permissions Icon Permissions

While seizure activity may be electrographically widespread, increasing evidence has suggested that ictal discharges may in fact represent travelling waves propagated from a focal seizure source. Interictal epileptiform discharges (IEDs) are an electrographic manifestation of excessive hypersynchronization of cortical activity that occur between seizures and are considered a marker of potentially epileptogenic tissue. The precise relationship between brain regions demonstrating IEDs and those involved in seizure onset, however, remains poorly understood. Here, we hypothesize that IEDs likewise reflect the receipt of travelling waves propagated from the same regions which give rise to seizures.

Forty patients from our institution who underwent invasive monitoring for epilepsy, proceeded to surgery and had at least one year of follow-up were included in our study. Interictal epileptiform discharges were detected using custom software, validated by a clinical epileptologist.

We show that IEDs reach electrodes in sequences with a consistent temporal ordering, and this ordering matches the timing of receipt of ictal discharges, suggesting that both types of discharges spread as travelling waves. We use a novel approach for localization of ictal discharges, in which time differences of discharge receipt at nearby electrodes are used to compute source location; similar algorithms have been used in acoustics and geophysics. We find that interictal discharges co-localize with ictal discharges. Moreover, interictal discharges tend to localize to the resection territory in patients with good surgical outcome and outside of the resection territory in patients with poor outcome. The seizure source may originate at, and also travel to, spatially distinct IED foci.

Our data provide evidence that interictal discharges may represent travelling waves of pathological activity that are similar to their ictal counterparts, and that both ictal and interictal discharges emerge from common epileptogenic brain regions. Our findings have important clinical implications, as they suggest that seizure source localizations may be derived from interictal discharges, which are much more frequent than seizures.

In focal epilepsy, seizures tend to be stereotyped within a given patient, with relatively consistent patterns of hypersynchronous activity observed in EEG recordings. 1 , 2 The concept of the seizure focus arose from these observations, in concert with the ability of focal surgical resections to lead to seizure freedom. Visual identification of the seizure onset zone through intracranial EEG (iEEG) monitoring is currently the gold standard for approximating the source of epileptiform activity, particularly in non-lesional patients. Surgical resections guided by such clinical inspections of the iEEG recordings have offered patients with drug-resistant focal epilepsy improved chances for seizure freedom.

Despite the relative successes of this approach, however, recent studies based on microelectrode recordings have provided evidence that ictal discharges seen in macroelectrode recordings may not necessarily indicate brain regions that are actively seizing. The recorded discharges may instead reflect receipt of activity from a seizure focus either in the form of travelling waves along the cortical surface or through white matter propagation. 3-10 Rapid propagation of ictal discharges and slow propagation of the sources of this activity appear to be relatively stereotyped within a given patient across seizures, suggesting preferred propagation pathways that are likely driven by a combination of proximity to the source of the activity, functional connections and susceptibility to recruitment. 11-13 These observations together suggest that the source of epileptiform activity may be much more spatially circumscribed than the observed seizure onset zone.

Interictal epileptiform discharges (IEDs) are another electrographic manifestation of excessive hypersynchronization of cortical activity that occur between seizures and are considered a marker of potentially epileptogenic tissue. 14 Even within individual patients, IEDs demonstrate significant spatial and temporal variability, likely related to changes in local and global brain states including state of arousal or temporal proximity to seizures. 14-18 Despite this variability, IEDs tend to involve a relatively consistent spatial core region, suggesting that the observed interictal discharges may also represent receipt of signal from a focal source. 14 , 19-22 Indeed, recent research has suggested that interictal discharges may also spread as travelling waves, much like ictal discharges. 10 The precise relationship between brain regions demonstrating IEDs, those involved in seizure onset and subsequent propagation and the potential sources of this activity, however, remains poorly understood. 14 , 23

Here, we hypothesize that focal sources of epileptiform activity emit travelling waves that underlie both interictal and ictal activity observed in iEEG recordings. We reasoned that if both ictal and interictal discharges reflect the receipt of travelling waves emitted from the same focal source, we should expect similar timings of discharges in the interictal and ictal data. We show that the temporal order of receipt of interictal discharges across sequences of electrodes is relatively consistent over time and mirrors the order observed during ictal discharges. We then use the timing of interictal discharge receipt to localize the source of epileptiform activity. In previous work, we used phase differences of ictal discharges at adjacent electrodes to localize the source of seizure activity, and we adopt a similar approach here. 9 Similar algorithms have been used extensively for signal detection in acoustics and radar 24-28 and for earthquake epicentre detection in geophysics. 29-33 We show that the source of IEDs, when determined in this way, co-localize with the source of seizure activity and with the resection territory in patients with good outcome.

Participants

Participants underwent a surgical procedure for placement of intracranial electrodes followed by iEEG monitoring in anticipation of a surgical resection for drug-resistant epilepsy. We performed all surgical procedures and iEEG monitoring at the Clinical Center at the National Institutes of Health (NIH; Bethesda, MD). In each case, the clinical team determined the placement of the contacts to localize epileptogenic regions. Outcome was evaluated using the Engel class. 34 Here, Engel class 1 reflects freedom from disabling seizures. Engel class 2 reflects rare disabling seizures. Engel class 3 reflects a worthwhile improvement relative to prior to surgery. Finally, Engel class 4 reflects no worthwhile improvement compared to prior to surgery. All patients had at least 1 year of follow-up. Engel classes were assessed at two years or time of last follow-up, whichever was earlier. Mean time from surgery at outcomes assessment was 20.73 ± 5.87 months. Engel class outcomes and follow-up times are provided in Table 1 . The research protocol was approved by the Institutional Review Board, and informed consent was obtained from the participants. Data were analysed using custom MATLAB scripts (Mathworks, Natick, MA). All data are reported as mean ± standard deviation unless otherwise reported.

Patient demographic information

ATL = anterior temporal lobectomy; B = bilateral; F = frontal; F2BTCS = focal to bilateral tonic–clonic seizures; FCD = focal cortical dysplasia; FIAS = focal impaired awareness seizure; HS = hippocampal sclerosis; L = left; LGG = low-grade glioma; MDG = microdysgenesis; MTS = mesial temporal sclerosis; NL = no lesion; P = parietal; PVNH = periventricular nodular heterotopia; R = right; SAH = selective amygdalo-hippocampectomy; T = temporal.

IED detector and IED sequences

For detection of interictal epileptiform discharges, we used a custom-built IED detector, used in previous research and validated by a clinical epileptologist. 18 The IED detector works by detecting large fluctuations in the iEEG trace voltage. 35 , 36 For each electrode, we z -scored the iEEG traces over the duration of the 2-h epoch. We then identified negative deflections in each iEEG trace with a prominence of at least 3σ and with width less than or equal to 50 ms. When such a deflection was discovered, we searched for an additional positive deflection with a peak prominence of at least 3σ that occurs within 100 ms of the negative peak, either before or after. We additionally required that the difference between positive and negative peak height was at least 9σ. The time of the negative peak was marked for further analysis.

We defined an IED sequence as an event in which IEDs were detected in three distinct electrodes within the same 100 ms window. 37-39 Additional details are provided in the Supplementary material.

Source localization

We hypothesized that there exists one or multiple focal sources which give rise to pathologic activity, including interictal as well as ictal discharges. IEDs may thus represent a single discharge that is released from a focal source and that radially spreads outward over the surface of the cerebral cortex over macroscopic-scale distances. 7-10 These waves thus reach different electrodes at different times. With this conceptualization, we should then be able to use the time of receipt of IEDs at different electrodes to estimate the location of the source of the pathological waves. Similar approaches have been used extensively in geophysics, acoustics and radar, and are referred to as multilateration algorithms. 24-33

Consider a source, s , which releases a signal that spreads outward at speed c , and reaches two electrodes, e i and e j . The signal reaches e i with the time delay τ i and e j with the time delay τ j . We are therefore able to express the difference in distance between s and e i , on the one hand, and s and e j , on the other, as:

The relative distance between the source and the two electrodes can therefore be expressed using the difference in time delays τ i − τ j required for the signal to propagate from the source to the two electrodes. 28 The estimate d s , e i − d s , e j can then be used to constrain the location of the source to a hyperbola over the brain surface. We model the brain surface for each patient as a geodesic mesh based on preoperative MRI. With multiple participating electrode pairs and multiple estimates of relative distance, we are able to create multiple hyperbolae, each of which contains the putative source. We then solve for the source location as the intersection of these hyperbolae (see Supplementary material ).

Our algorithm assumes that the signal spreads outward evenly, in a concentric circular fashion, over the grey matter. This assumption is challenged in a simulation analysis in Supplementary Fig. 4 .

Data and code availability

Data not provided in the article because of space limitations may be shared (anonymized) at the request of any qualified investigator for purposes of replicating procedures and results.

Our custom scripts used for IED detection and source localization are publicly available for download at https://research.ninds.nih.gov/zaghloul-lab/downloads .

We examined iEEG recordings in 40 participants (35.40 ± 10.20 years old; 18 females) with drug-resistant epilepsy who were monitored for seizures with subdural and/or depth electrodes ( Table 1 ). All patients had focal epilepsy and had seizures during the recording period. All patients proceeded to surgery for surgical resection of the brain regions thought to give rise to seizures and had at least one year of clinical follow-up. Seizure data were inaccessible in four patients due to the files becoming lost in our database; these patients were excluded from analyses which required the presence of seizure data. Engel class outcomes were assessed at 2 years (see ‘Materials and methods’ section). Of the 40 total participants, 19 achieved an Engel class 1 outcome following surgical resection, 9 achieved an Engel class 2 outcome, 8 achieved an Engel class 3 outcome and 4 achieved an Engel class 4 outcome. Mean follow-up was 20.73 ± 5.87 months. Engel class outcomes and follow-up times are provided in Table 1 . For one patient (Patient 27), outcome changed from Engel class 3a to 1c at 60 months. This patient’s 16-month outcome, 3a, was considered in the analysis. No other patient went from non-seizure-free to seizure-free, or vice versa, between the 2-year outcome and long-term outcome. For the purposes of analysis, we divided our participants into those with good 2-year post-surgical seizure outcome (Engel 1, freedom from disabling seizures) and those with poor post-surgical seizure outcome (Engel 2 or greater, disabling seizures persist).

Interictal discharge sequences are stereotyped

We were interested in examining and understanding the origin of sequences of IEDs detected by the recording electrodes. We hypothesized that IEDs emerge from focal brain regions that emit travelling waves, which spread outward in a radial fashion. If this were the case, then we would expect to observe sequences of IEDs in recording electrodes that exhibit a consistent and predictable ordering.

To test this hypothesis, we used an automated IED detector to identify all sequences of IEDs recorded in the implanted electrodes (Fig. 1A and B; see ‘Material and methods’ section). We retained all sequences (or subsequences; see Supplementary material ) of exactly three IEDs that occurred within a single 100 ms window (see ‘Materials and methods’ section). In an example participant, we observed many different IED sequences ( Fig. 1C ). Each sequence is defined by a specific order of activity across three electrodes, and we can quantify the probability of observing each specific three-member sequence in every participant. We observed that IED sequences involving a set of electrodes often exhibit a predictable and consistent temporal ordering, and that it is unusual to find the same electrodes involved in a sequence with a differing order.

Interictal discharge sequences are stereotyped. (A) Implanted subdural and depth electrodes are shown on a cortical surface reconstruction in a single representative participant. (B) An example discharge is shown, detected in three electrodes in the anterolateral temporal lobe (ALT). (C) The most commonly occurring IED sequences involving three electrodes are shown for this participant. Each sequence begins with an IED observed in the first electrode (inner ring), followed by the second and third electrode, shown in the second and third rings respectively. Arc length of each ring represents the probability of observing an IED in that electrode in the sequence. Only the 10 electrodes that are most frequently observed to lead an IED sequence are represented in the first ring. Similarly, the second and third rings only display the 10 most frequently involved electrodes in the second and third position of the IED sequences, given the IED observed in the first electrode. Electrodes are coloured based on their membership in each subdural grid or strip. (D) Likelihood of observing one of the 20 most commonly occurring IED sequences across participants (black, mean likelihood 4.8 × 10–03 ± 4.9 × 10–03). Likelihood of observing IED sequences in the same electrodes but in a different order is significantly less across participants (red, shuffled, 1.6 × 10–03 ± 1.9 × 10–03). Observing IED sequences in randomly selected sets of three electrodes across participants are even less likely (grey, random, 3.1 × 10–04 ± 2.9 × 10–04). Error bars reflect standard error of the mean.

Interictal discharge sequences are stereotyped. ( A ) Implanted subdural and depth electrodes are shown on a cortical surface reconstruction in a single representative participant. ( B ) An example discharge is shown, detected in three electrodes in the anterolateral temporal lobe (ALT). ( C ) The most commonly occurring IED sequences involving three electrodes are shown for this participant. Each sequence begins with an IED observed in the first electrode (inner ring), followed by the second and third electrode, shown in the second and third rings respectively. Arc length of each ring represents the probability of observing an IED in that electrode in the sequence. Only the 10 electrodes that are most frequently observed to lead an IED sequence are represented in the first ring. Similarly, the second and third rings only display the 10 most frequently involved electrodes in the second and third position of the IED sequences, given the IED observed in the first electrode. Electrodes are coloured based on their membership in each subdural grid or strip. ( D ) Likelihood of observing one of the 20 most commonly occurring IED sequences across participants (black, mean likelihood 4.8 × 10 – 03 ± 4.9 × 10 – 03 ). Likelihood of observing IED sequences in the same electrodes but in a different order is significantly less across participants (red, shuffled, 1.6 × 10 – 03 ± 1.9 × 10 – 03 ). Observing IED sequences in randomly selected sets of three electrodes across participants are even less likely (grey, random, 3.1 × 10 – 04 ± 2.9 × 10 – 04 ). Error bars reflect standard error of the mean.

In order to quantify the extent to which the temporal orderings of the IED sequences are predictable, we identified the 20 most commonly observed three-member sequences in each participant. The likelihood of each of these sequences was computed, by taking the count and dividing by the total number of three-member sequences in each patient. Mean likelihood for the top 20 sequences was 4.8 × 10 – 03 ± 4.9 × 10 – 03 . This analysis was then repeated after permuting the order of the top 20 sequences. For permuted top sequences, mean likelihood was 1.6 × 10 – 03 ± 1.9 × 10 – 03 . Finally, it was repeated after retrieving 100 random sequences of three electrodes. We only chose sets of electrodes that were located within 30 mm of one another, to address the possible confound in which, in electrodes that are far from each other, sequences might be rare simply by virtue of the electrode spacing. Mean likelihood for random sequences was 3.1 × 10 – 04 ± 2.9 × 10 –04 . There was a significant difference between at least two groups [one-way ANOVA, F (2,117) = 23.76, P < 0.0001]. Tukey’s Honest Significant Difference (HSD) test for multiple comparisons revealed significant differences between the true top 20 sequences and their permuted counterparts ( P < 0.0001) and between the top 20 sequences and the random sequences ( P < 0.0001), but not between the permuted and random sequences ( P = 0.15). These results are illustrated in Fig. 1D . Here, in each patient, likelihoods of sequences were binned, with the same binning scheme used in all patients. Bin counts were then averaged across patients. These data demonstrate that the observed IED sequences occur more frequently than the permuted or random sequences. This suggests that the observed IED sequences may in fact arise from focal regions of pathological tissue, and therefore arrive at the recording electrodes with reliable orderings.

Temporal order of IED sequences is preserved in the ictal data

We wished to test the hypothesis that ictal and interictal discharges arise from similar regions. 10 If this were true, then we would expect that the temporal ordering of interictal discharges should be similar to that of the discharges observed during seizures.

In the same example participant as discussed in Fig. 1A–C , we identified two commonly occurring IED sequences. In the first sequence, the electrodes involved reside in the right parietal lobe, while in the other, the electrodes are in the right anterior temporal lobe. We then examined the time series captured from these same electrodes during a typical seizure ( Fig. 2A ). Earlier during the seizure, ictal discharges are present in the parietal electrodes but not in the temporal electrodes. Later, ictal discharges are present in the temporal lobe but not in the parietal lobe. Close examination in one-second epochs reveals that the ictal discharges appear to arrive with a reliable ordering that matches the order observed in the respective IED sequences. The change in ictal discharge patterns suggests that the source of seizure activity may migrate between distinct foci. 9 Moreover, at each focus, it appears that the source of seizure activity emits discharges that match the order of observed IED sequences.

We directly compared the order of discharges in the ictal and interictal data. To do so, we first chose one particular common IED sequence of three members. We then retrieved every IED sequence in this patient involving the same three electrodes, but in any ordering. We then extracted the latency between IED discharges observed in the first and second electrode members of the common IED sequence, and between the second and third members. Because any ordering of the sequence is permitted, latencies can be positive or negative. In the two example sets of electrodes from the same participant, most combinations of IED latencies arise in the bottom left quadrant, suggesting that both latencies are negative. In other words, most IED latencies exhibit a temporal order similar to the commonly observed IED sequence, although all combinations of latencies across the set of three electrodes are also seen ( Fig. 2B ). We then examined sequences of ictal discharges detected in the same sets of electrodes. Similarly to the IED sequences, we considered ictal discharges involving the same electrodes, but in any order, and extracted the latencies between constituent members. The ictal discharges appear to exhibit a temporal ordering that is similar to that of the interictal discharges.

We repeated this analysis for the 20 most common three-member IED sequences observed in this participant. For each sequence, we took all sequences involving those electrodes, but in any order, and retrieved the latencies between discharges in the first and second members of the common sequence, and between the second and third members. We aggregated these latencies across all 20 IED sequences. As expected, most observed latencies exhibit a temporal ordering similar to the commonly observed IED sequences ( Fig. 2C , left). We then examined ictal discharges involving these same sets of electrodes. The ictal discharges also exhibit a similar combination of latencies across the involved electrodes ( Fig. 2C , right). We found a similar correspondence across all participants ( Fig. 2D ), suggesting that the temporal order of discharges during seizures appear to match the temporal order observed during IED sequences.

We wished to determine whether the propensity of observed discharges to match the ordering of common discharges was greater than expected by chance. To study this, we only considered latencies less than or equal to 20 s, to reduce the influence of noise. For latencies less than or equal to 20, the weighted centroid of the heatmap of IED latencies ( Fig. 2D , left) was −1.92, −2.42 s, for the first and second latencies, respectively. We then divided latencies into quadrants, and asked if the bottom left quadrant had the greatest mean count. Mean count in the bottom left, bottom right, top left and top right quadrants were 10.08 ± 9.69, 5.06 ± 4.70, 5.45 ± 4.45 and 2.39 ± 3.58, respectively. There was a significant difference between at least two groups [one-way ANOVA, F (3,480) = 33.17, P < 0.0001]. Tukey’s HSD test for multiple comparisons revealed significant differences between the bottom left quadrant and all other quadrants ( P < 0.0001). Next, we considered the ictal discharges. For latencies less than or equal to 20, the weighted centroid of the heatmap of ictal latencies ( Fig. 2D , right) was −0.87, −0.81 s. We again divided latencies into quadrants. Mean count in the bottom left, bottom right, top left and top right quadrants were 4.37 ± 4.54, 3.14 ± 2.10, 3.18 ± 2.01 and 2.11 ± 1.93, respectively. There was a significant difference between at least two groups [one-way ANOVA, F (3,480) = 12.61, P < 0.0001]. Tukey’s HSD test for multiple comparisons revealed significant differences between the bottom left quadrant and the bottom right quadrant ( P = 0.005), the top left quadrant ( P < 0.007) and the top right quadrant ( P < 0.001).

Interictal and ictal discharges localize to similar regions

The reliability of the IED sequences, and the preserved temporal order seen in the ictal data, suggest that the involved electrodes may receive pathological waves of activity from a particular focal source. In this conceptualization, a source of epileptiform activity emits a signal that travels radially outward over the cerebral cortex and reaches different electrodes at different times, depending on their distance from the source (Fig. 3A and B). 9 We can therefore use the time differences observed across multiple pairs of electrodes to estimate the location of this potential source ( Fig. 3C ; see ‘Materials and methods' section). We were interested in comparing the results from source localization using the interictal discharges versus using the ictal data.

Temporal order of IED sequences is preserved in the ictal data. (A) Ictal time series for two sets of three electrodes in a single seizure in an example participant. Each set of three electrodes represents a common IED sequence. One set is in the parietal lobe (PG) and the other is in the temporal lobe (ALT). Left: cortical reconstruction of the electrode locations. Two magnified 1-s epochs (bottom) are shown, captured at 23–24 s and 53–54 s, respectively, following seizure onset (top). Discharges from each set of electrodes arrive in a consistent temporal order. (B) The latencies of all observed IED sequences involving the same parietal and temporal sets of three electrodes. Latencies are arranged by the difference in time between an IED appearing on the first and second electrode members of the commonly observed IED sequence, and between the second and third member (black rings). Points in the bottom left quadrant therefore represent a temporal order identical to the commonly observed IED sequence in those electrodes. Latencies in the ictal discharges observed in the same electrodes during seizures demonstrate a similar order as the common IED sequences (filled circles, colour-coded by duration after seizure onset). (C) Number of IED sequences, and their relative latencies, observed in all sets of three electrodes that were involved in the 20 most commonly observed IED sequences in this participant (left). Number and relative latencies of ictal discharges observed in the same sets of electrodes (right). (D) Mean number of IED sequences (left) and ictal discharges (right), and their relative latencies, observed in each of the 20 most commonly observed IED sequences per patient, aggregated across all participants.

Temporal order of IED sequences is preserved in the ictal data. ( A ) Ictal time series for two sets of three electrodes in a single seizure in an example participant. Each set of three electrodes represents a common IED sequence. One set is in the parietal lobe (PG) and the other is in the temporal lobe (ALT). Left : cortical reconstruction of the electrode locations. Two magnified 1-s epochs ( bottom ) are shown, captured at 23–24 s and 53–54 s, respectively, following seizure onset ( top ). Discharges from each set of electrodes arrive in a consistent temporal order. ( B ) The latencies of all observed IED sequences involving the same parietal and temporal sets of three electrodes. Latencies are arranged by the difference in time between an IED appearing on the first and second electrode members of the commonly observed IED sequence, and between the second and third member (black rings). Points in the bottom left quadrant therefore represent a temporal order identical to the commonly observed IED sequence in those electrodes. Latencies in the ictal discharges observed in the same electrodes during seizures demonstrate a similar order as the common IED sequences (filled circles, colour-coded by duration after seizure onset). ( C ) Number of IED sequences, and their relative latencies, observed in all sets of three electrodes that were involved in the 20 most commonly observed IED sequences in this participant ( left ). Number and relative latencies of ictal discharges observed in the same sets of electrodes ( right ). ( D ) Mean number of IED sequences ( left ) and ictal discharges ( right ), and their relative latencies, observed in each of the 20 most commonly observed IED sequences per patient, aggregated across all participants.

Latencies between interictal and ictal discharges can be used to estimate the location of a hypothesized source of activity. (A) We hypothesize that one or multiple focal sources (star) emit pathological waves of activity that spread radially outward over the brain surface. Discharges are detected on electrodes with different latencies; these latencies are used to estimate Δd, the relative difference in distance between the electrodes and the hypothesized source. For each pair of electrodes, this estimate constrains the location of the hypothesized source to a single hyperbola. With multiple electrode pairs, we estimate the location of the source location from the intersection of the hyperbolae. (B) A representative IED sequence in a set of electrodes in the anterior temporal lobe. The latencies between each IED, Δt, can be used to estimate Δd. (C) The estimates of Δd for each electrode pair constrain the source to a hyperbola in geodesic space. We compute the source of these travelling waves (star) as the intersection of these hyperbolae. (D) Left: All IEDs are detected in a set of three electrodes in the parietal lobe (top) and another set of three electrodes in the anterior temporal lobe (bottom). We estimate the location of the hypothesized source for every IED sequence involving these electrodes and plot the number of times the estimated source location falls within each region of interest as a heatmap. Right: All ictal discharges during seizures are detected in the same sets of electrodes. Using the phase differences of ictal discharges across electrodes, we estimate the location of the hypothesized source of ictal discharges. Colour intensity indicates the number of times an estimated source of these ictal discharges localizes to each region of interest in these brain regions and colour hue indicates the mean time from seizure onset at which the source localization occurred. For each set of three electrodes, discharges tend to localize to similar regions in interictal versus ictal states.

Latencies between interictal and ictal discharges can be used to estimate the location of a hypothesized source of activity. ( A ) We hypothesize that one or multiple focal sources (star) emit pathological waves of activity that spread radially outward over the brain surface. Discharges are detected on electrodes with different latencies; these latencies are used to estimate Δ d , the relative difference in distance between the electrodes and the hypothesized source. For each pair of electrodes, this estimate constrains the location of the hypothesized source to a single hyperbola. With multiple electrode pairs, we estimate the location of the source location from the intersection of the hyperbolae. ( B ) A representative IED sequence in a set of electrodes in the anterior temporal lobe. The latencies between each IED, Δ t , can be used to estimate Δ d . ( C ) The estimates of Δ d for each electrode pair constrain the source to a hyperbola in geodesic space. We compute the source of these travelling waves (star) as the intersection of these hyperbolae. ( D ) Left : All IEDs are detected in a set of three electrodes in the parietal lobe ( top ) and another set of three electrodes in the anterior temporal lobe ( bottom ). We estimate the location of the hypothesized source for every IED sequence involving these electrodes and plot the number of times the estimated source location falls within each region of interest as a heatmap. Right : All ictal discharges during seizures are detected in the same sets of electrodes. Using the phase differences of ictal discharges across electrodes, we estimate the location of the hypothesized source of ictal discharges. Colour intensity indicates the number of times an estimated source of these ictal discharges localizes to each region of interest in these brain regions and colour hue indicates the mean time from seizure onset at which the source localization occurred. For each set of three electrodes, discharges tend to localize to similar regions in interictal versus ictal states.

We examined the same two commonly occurring IED sequences in the same example participant. We identified all IED sequences involving those three electrodes, but in any order. Based on those sequences, we estimated the location of the source of IED activity for each sequence. We mapped localized points to regions of interest that were evenly spaced across the cortical surface (see Supplementary material ). Results are plotted on a heatmap on a cortical surface reconstruction, where warmer colours designate greater source localizations per region of interest ( Fig. 3D , left). We then examined the discharges captured during a single representative seizure in these two sets of electrodes. This was the same seizure as was considered in Fig. 2 . Results are again plotted on a cortical surface reconstruction ( Fig. 3D , right). Here, hue designates time from seizure onset (blue is early; yellow is late). Meanwhile, intense colours designate a greater number of source localizations per region of interest. In both sets of electrodes, localization based on the ictal discharges is similar to localization based on the IED sequences, as would be expected given the similar temporal ordering of discharges observed in the ictal and interictal data. For the parietal leads, IEDs localize to a gyrus in the inferior parietal lobule, with the strongest region of interest posterolateral to the leads. The seizure also localizes to the inferior parietal lobule, here with the strongest region of interest posteromedial to the leads. In this analysis, the distance between strongest seizure region of interest and strongest IED region of interest is 20.5 mm. However, of note, there does appear to be a secondary IED focus, which is posteromedial to the parietal leads and which coincides spatially with the strongest seizure focus. For the temporal leads, both IEDs and seizure discharges localize to the temporal pole, with strongest activity in the inferior temporal gyrus, parahippocampal gyrus and middle temporal sulcus. In this analysis, the strongest IED region of interest is the same as the strongest seizure region of interest.

We next considered all electrodes in this patient, and examined all captured IED sequences, as well as all ictal discharges captured during the same representative seizure. In this analysis, we retained all full IED sequences of length at least three, rather than length exactly three, recorded in all electrodes (see Supplementary material ). Using these sequences and our source localization procedure, we identified the brain regions most commonly identified as the estimated source of the observed IED sequences ( Fig. 4A , left; see ‘Materials and methods' section). We then estimated the source of ictal data at every time point in the same representative seizure, based on the relative phase of ictal discharges observed across all recording electrodes. Localization based on the ictal data during this seizure is visually similar to that based on the IED sequences in this participant ( Fig. 4A , right). The seizure source appears to originate from and travel to distinct IED foci (see Supplementary Fig. 2A ). This patient went on to have a right parietal lesionectomy (grey region). Outcome was Engel class 3a at 16 months. Strongest points of IED and seizure localization were both outside of the resection territory, suggesting a successful prediction of our algorithm (see Supplementary Fig. 1 , Patient 27).

Interictal and ictal discharges localize to similar brain regions. (A) Left: Estimated source location for all IED sequences captured in all electrodes in an example participant. Each IED sequence generates an estimated source location. Colour and size of points indicate the number of times a source was estimated to localize within 2 mm of that location across all observed IED sequences. Estimated locations are collapsed to two dimensions. Resection cavity indicated in grey. Right: Estimated source location for all ictal discharges captured in all electrodes during a seizure in the same participant. The size of each point indicates the number of times a source was estimated to localize within 2 mm of that location. Colour indicates time from seizure onset. (B) Correlation between the number of IED sequences that localized to each region of interest and the number of ictal discharges that localized to the same region of interest (Spearman’s ρ = 0.64). Each point represents one region of interest and is coloured based on the mean time from seizure onset when ictal discharges are localized to that region of interest. Axes are logarithmically scaled and normalized to the number of times discharges were observed in the region of interest containing the most localizations. (C) We compared the true correlation coefficient with the correlations observed by chance, generated by shuffling the lead labels 100 times. (D) We performed a similar shuffling procedure in each participant, comparing the true correlation coefficient (black) with the median and interquartile range of correlation coefficients generated by chance after shuffling lead labels 100 times in each participant (red). (E) Left: In every participant, we computed the Euclidian distance between the region of interest most frequently containing the estimated source location, based on the IED sequences and the region of interest most frequently containing the source location based on ictal discharges in the true data (black, histogram with 10 mm bins). We similarly computed distances between estimated IED and ictal source locations after shuffling lead labels 100 times in each participant (red). Right: The most frequently estimated location of the source of IED sequences falls within 10 mm of the most frequently estimated location of the source of ictal discharges in nine participants, significantly greater than the number observed by chance after shuffling the lead labels.

Interictal and ictal discharges localize to similar brain regions. ( A ) Left : Estimated source location for all IED sequences captured in all electrodes in an example participant. Each IED sequence generates an estimated source location. Colour and size of points indicate the number of times a source was estimated to localize within 2 mm of that location across all observed IED sequences. Estimated locations are collapsed to two dimensions. Resection cavity indicated in grey . Right : Estimated source location for all ictal discharges captured in all electrodes during a seizure in the same participant. The size of each point indicates the number of times a source was estimated to localize within 2 mm of that location. Colour indicates time from seizure onset. ( B ) Correlation between the number of IED sequences that localized to each region of interest and the number of ictal discharges that localized to the same region of interest (Spearman’s ρ = 0.64). Each point represents one region of interest and is coloured based on the mean time from seizure onset when ictal discharges are localized to that region of interest. Axes are logarithmically scaled and normalized to the number of times discharges were observed in the region of interest containing the most localizations. ( C ) We compared the true correlation coefficient with the correlations observed by chance, generated by shuffling the lead labels 100 times. ( D ) We performed a similar shuffling procedure in each participant, comparing the true correlation coefficient (black) with the median and interquartile range of correlation coefficients generated by chance after shuffling lead labels 100 times in each participant ( red ). ( E ) Left : In every participant, we computed the Euclidian distance between the region of interest most frequently containing the estimated source location, based on the IED sequences and the region of interest most frequently containing the source location based on ictal discharges in the true data (black, histogram with 10 mm bins). We similarly computed distances between estimated IED and ictal source locations after shuffling lead labels 100 times in each participant (red). Right : The most frequently estimated location of the source of IED sequences falls within 10 mm of the most frequently estimated location of the source of ictal discharges in nine participants, significantly greater than the number observed by chance after shuffling the lead labels.

We were interested in more rigorously capturing the co-localization of interictal and ictal discharges. Therefore, we performed a similar localization procedure for all seizures recorded in this participant (eight seizures) and compared the results to the localization obtained from the IED sequences. Both ictal and interictal localized points were mapped to brain surface regions of interest. We found a strong correlation between the number of IED sequences that localized to each region of interest and the number of ictal discharge localizations to the same region of interest (Spearman’s ρ = 0.64; Fig. 4B ). To confirm that this correspondence in localization between ictal and interictal activity is greater than chance while adjusting for confounding factors such as electrode placement, we used a shuffling procedure. During each of 100 iterations, we shuffled the electrode labels and recomputed IED sequences and IED localization. In other words, the mapping between raw time series and spatial location was scrambled over all electrodes in each patient’s implant. Shuffling was done prior to assembly of IEDs into sequences, so that electrodes did not need to record IEDs in order to be included in the shuffled sequences. An additional shuffling analysis, in which lead labels were shuffled after assembly into IED sequences, is considered in the legend of Supplementary Fig. 3 . Ictal localization was left unchanged. The true value of ρ is in the 94th percentile of shuffled values ( Fig. 4C ).

We repeated this analysis in all participants and found that, in the vast majority of participants, the true correlation between IED and ictal localization is significantly greater than the correlation that would emerge by chance ( Fig. 4D ). In every participant, we additionally selected the region of interest that was most frequently identified as the source of interictal discharges, on the one hand, and of ictal discharges, on the other (see Methods). We then took the Euclidean distance between these two regions in each patient. Across participants, distance between ictal and ictal discharge sources is significantly less than that which would arise by chance, after shuffling the electrode labels [28.68 ± 20.75 mm versus 42.90 ± 21.97 mm; t (57) = 2.51, P = 0 .01; Fig. 4E , left]. The most frequently estimated location of the source of IED sequences falls within 10 mm of the putative ictal source in nine participants, significantly greater than the number observed by chance after shuffling the electrode labels (3.70 ± 1.66, true value is in 100th percentile; Fig. 4E , right).

Note that the findings described in Fig. 4D would be expected given the results shown in Fig. 2D . However, the analysis shown in Fig. 4D provides additional information for several reasons. First, all electrodes and sequences are considered in Fig. 4D , rather than just the top 20 most common three-member sequences. Second, while in Fig. 2D we showed that ictal discharge latencies reside in the bottom left quadrant, the findings in Fig. 4D rest not just upon the fact that the signs of the latencies in the ictal data match those in the interictal data, but, additionally, that the values themselves are similar. Finally, in Fig. 4D , the temporal data provided in Fig. 2D are subject to spatial constraints, namely, the brain anatomy of each individual patient.

The hypothesized source of interictal and ictal discharges lies within the resection territory in participants with good outcome

Our data suggest a putative focal source of IEDs may account for the discharge sequences observed across the recording electrodes. It is challenging, however, to definitively establish which brain regions are indeed responsible for ictal and interictal epileptiform activity. A common approach to address this challenge is to determine whether the resection of a brain region leads to post-surgical seizure control. This would imply that epileptogenic tissues resided within the resection.

Therefore, in every participant, we identified the brain region of interest most frequently selected as the IED source. We similarly identified the region of interest most frequently selected as the ictal source. In participants with good surgical outcomes, the majority of IED and ictal localizations fell within the resection cavity ( Fig. 5A ). On the other hand, in participants with poor surgical outcomes, the majority fell outside the resection cavity. These data demonstrate that localization based on either the IED sequences or the ictal discharges appears to identify brain regions that are involved in epileptogenic activity. This therefore suggests that a focal source of pathologic travelling waves could underlie the observed time differences across electrodes in the IED sequences and in the ictal discharges.

The hypothesized source of interictal and ictal discharges lies within the resection territory in participants with good outcomes. (A) In every participant with good or poor surgical outcomes, we determined whether the region most frequently identified as the source of IED sequences or ictal discharges lay within or outside the resection cavity. (B) We compared the number of times in which source localization based on the IED sequences fell within the resection cavity, in participants with good surgical outcomes (13 participants out of 19 total, true data indicated by blue line), to the number of times this would occur by chance after shuffling the lead labels 100 times in each participant (8.88 ± 1.61, blue histogram). We also compared the number of times in which source localization based on the IED sequences identified a region that fell outside of the resection cavity in participants with poor surgical outcomes (17 participants out of 21 total, true data indicated with yellow line), to the number of times this would occur by chance after shuffling the lead labels 100 times in each participant (15.39 ± 1.72, yellow histogram). In this analysis, sensitivity corresponds to the number of participants with good surgical outcomes where localization falls within the resection cavity. Specificity corresponds to the number of participants with poor outcomes where localization falls outside the resection cavity.

The hypothesized source of interictal and ictal discharges lies within the resection territory in participants with good outcomes. ( A ) In every participant with good or poor surgical outcomes, we determined whether the region most frequently identified as the source of IED sequences or ictal discharges lay within or outside the resection cavity. ( B ) We compared the number of times in which source localization based on the IED sequences fell within the resection cavity, in participants with good surgical outcomes (13 participants out of 19 total, true data indicated by blue line), to the number of times this would occur by chance after shuffling the lead labels 100 times in each participant (8.88 ± 1.61, blue histogram). We also compared the number of times in which source localization based on the IED sequences identified a region that fell outside of the resection cavity in participants with poor surgical outcomes (17 participants out of 21 total, true data indicated with yellow line), to the number of times this would occur by chance after shuffling the lead labels 100 times in each participant (15.39 ± 1.72, yellow histogram). In this analysis, sensitivity corresponds to the number of participants with good surgical outcomes where localization falls within the resection cavity. Specificity corresponds to the number of participants with poor outcomes where localization falls outside the resection cavity.

We compared the results of localization with respect to the resection territory, using the IED sequences, to that obtained from chance. To do so, we shuffled the electrode labels 100 times in each participant and identified the region of interest most frequently identified as the source of the IED sequences in each shuffle ( Fig. 5B ). In participants with good surgical outcomes, the number of patients, on average, with greatest region of interest within the resection territory was 8.88 ± 1.61, compared to 13 of 19 patients in the true data. The true value is in the 100th percentile of shuffled values. On the other hand, in patients with poor outcome, the number of patients, on average, with greatest region of interest outside of the resection territory was 15.39 ± 1.72, compared to 17 of 21 total patients in the true data. The true value is in the 74th percentile of the shuffled data. Therefore, while in our results specificity is greater than sensitivity, on the other hand, the true positive rate is significantly greater than that expected by chance, while this is not true of the true negative rate.

We were interested in comparing the results of our IED source localization method to methods which are simpler to implement. Therefore, we considered two controls. First, we considered the electrode that most commonly led IED sequences. Second, we considered the electrode with the most frequent IED activity. Our technique outperformed both simpler methods ( Supplementary Fig. 3A ). Of 19 patients with good post-surgical outcome, 8 had earliest electrode in the resection territory and 7 had most frequent electrode in the resection territory, compared to 13 with source in the resection in the true data. Among 21 patients with poor post-surgical outcome, 18 had earliest electrode outside of the resection territory and 17 had most frequent electrode outside the resection territory, compared to 17 with source outside of the resection in the true data. Overall accuracy was 65% in the earliest approach and 60% in the most frequent approach, compared to 75% under our approach. Clinical accuracy was 47.5%, in that 19 / 40 patients had Engel class 1 outcome. The putative IED location resulting from the use of simpler techniques tended to be relatively far from the source of IEDs localized using our computational approach. Across patients, the earliest electrode was 42.49 ± 42.18 mm from the localized source. The most frequent electrode was 50.81 ± 49.42 mm from the localized source ( Supplementary Fig. 3B ).

We were curious as to whether the early seizure source holds particular consequence for seizure freedom, as opposed to the greatest seizure source during all seizure activity. Therefore, we performed an additional analysis where only early seizure activity is considered (see Supplementary material and Supplementary Table 1 ). Prediction is more accurate in some patients under the early seizure analysis but less accurate in others, and overall, accuracy is the same under both techniques. The strongest IED focus may serve as a marker for the site of seizure activity with greatest consequence for seizure control, even in cases where there is a discrepancy between the early source location and the eventual source location.

Our results suggest that interictal epileptiform discharges may be generated by focal sources of pathological activity, which emit travelling waves that reach surrounding brain regions with consistent patterns. Interictal discharges arise as sequences, measured in sets of electrodes with consistent temporal orderings. Similar temporal orderings occur during ictal discharges, consistent with the hypothesis that both ictal and interictal discharges emerge from a common source. Using the times of receipt of the interictal discharges, we localized the source of interictal activity, and the location of this activity matches the results of ictal discharge localization. The majority of localizations lie within the resection territory in participants with good seizure outcomes and outside the resection in those with poor outcomes.

Visual inspection of the iEEG is currently the gold standard used to guide surgical resections in patients with drug-resistant focal epilepsy. Intracranial electrodes implanted for seizure monitoring, however, can only capture activity from a small fraction of the cortical volume, 40 and so much of the pathological activity recorded through these electrodes likely reflects propagated activity from nearby regions. Although the traditional interpretation of clinically observed ictal discharges in the electrode recordings has been that the underlying brain regions are actively seizing, studies using microelectrode recordings have challenged this interpretation, demonstrating that ictal discharges may in fact reflect receipt of travelling waves, propagated from a more spatially circumscribed seizure focus or ictal wavefront. 3-10 Our data provide evidence that interictal discharges may also reflect travelling waves of pathological activity, similar to their ictal counterparts. 3-10

Given the similar temporal ordering observed in both the interictal and ictal discharges, our results are consistent with recent evidence that the two types of activity may emerge from a common source. 10 The suggestion that interictal and ictal discharges are directly related is consistent with animal work showing propagation over functional pathways of both interictal and ictal discharges from a focal source, 12 and with simultaneous regional offsets of seizure activity presumably emanating from a common source. 13 In the clinical setting, brain regions with IED activity that leads other regions within the irritative zone appear to be correlated with the eventual site of seizure origin. 18 , 39 , 41-44 Patients with well-localized IEDs prior to surgery are more likely to have a favourable post-surgical outcome. 45 , 46 Finally, disappearance of IEDs postoperatively can be used to predict good seizure outcomes. 45-47 Notably, we did not find evidence of bidirectional interictal discharge propagation patterns that could correspond to a predominant as well as antipodal source, as has been previously reported. 10 This difference may be explained at least in part by differing methodologies, as bidirectional interictal discharges were reported only in IEDs measured from a microelectrode array that had been recruited by the ictal wavefront. Without microelectrode array data, we cannot formally establish whether electrodes are recruited into the ictal wavefront, and we may also lack the spatial resolution to detect bidirectional discharges.

Although our data suggest a strong correspondence between interictal and ictal discharges, the relation between brain regions involved in interictal epileptiform activity, referred to as the irritative zone, and those involved in seizure onset remains complex. There are several lines of evidence that suggest that the irritative zone is not monolithic. There are notable differences in the spatial extent and morphological appearance between interictal and ictal discharges, which could be due to variable propagation of this activity. 48 Similarly, variable spatial extent is often seen in interictal discharges over time even within the same patient, likely due to changes in local and global brain states such as state of arousal. 14-18 , 21 , 49-51 Of note, this variability in spread may exist even if the underlying source of IEDs remains the same. This observation in fact underscores a strength of the computational approach we use for discharge source localization. That is, given multiple IEDs from a single source, each of which spreads over the brain surface to a different extent, our algorithm is nonetheless capable of recovering the single source common to each of the discharges.

Apart from the variability induced by any one particular source, we found that the irritative zone often consists of multiple subpopulations of activity, sometimes nearby to each other and sometimes separated by lobes or hemispheres. In many patients, spontaneous seizures appear to arise only from one of these source regions (for example, see Supplementary Fig. 1 , Patients 17 and 27). The other independent IED populations perhaps represent less epileptogenic or only potentially epileptogenic brain regions that do not have the capacity to generate spontaneous seizures, but may be relatively more susceptible to recruitment as seizures evolve and spread. Our results are in agreement with recent work that found that there may be multiple co-incident seizure sources, some of which may move and others of which are stationary. 52 Spread of the seizure source to distinct IED foci may occur by direct, contiguous spread over grey matter or by distant travel via white matter pathways. We do in fact find that when brain regions involved in IEDs are recruited into seizure activity, whether at seizure onset or later in seizure evolution, the temporal ordering of the ictal discharges within that given brain region remains consistent with the interictal data. This suggests that irritable brain regions that are regional sources of IED activity may indeed have a lower seizure threshold and may be preferentially recruited as seizures evolve. The development of in vivo , widefield cell-type–specific calcium imaging has offered an opportunity to capture stereotypic patterns of ictal and interictal discharge origin and spread in model organisms, and represents a promising future avenue. 53-55

Using a novel computational approach based on the temporal orderings of discharges, we found that ictal and interictal discharges tend to localize to similar brain regions. Additionally, discharges tend to localize to the resection territory in patients with good seizure outcomes and outside of it in patients with poor outcomes. Source localization based on the interictal discharges is in fact slightly stronger at predicting outcomes in our data. One possible reason for this is that sources of seizure discharges may move over time, leading to more variable source localizations. On the other hand, discovered IED sources are presumably stationary. We explored whether resection of the early seizure source is of particular importance to seizure freedom, but found that focusing only on early seizure activity did not improve prediction accuracy ( Supplementary Table 1 ). It seems that both the site of seizure onset and of seizure travel may carry importance for seizure control. Resection of the site of seizure travel, for instance, may disconnect the seizure focus, rather than removing it, but may nonetheless provide seizure freedom. Finally, resection of the strongest focus of IED activity appears to be critical for seizure control, even in the setting of a moving seizure source, in which there is ambiguity in the seizure source location.

Our computational approach depends on the assumption that discharges arising from a pathological source spread outwardly over the grey matter at a fixed rate. Note that this limitation concerns the spread of discharges themselves, rather than the spread of the source of the discharges. We acknowledge that, rather than spreading evenly, discharges may instead spread over the grey matter but at variable rates or, alternatively, may spread over white matter tracts. Any non-local spread of discharges, whether over grey matter or white matter, may distort the results of our algorithm. Challenges to the assumptions that underlie our approach are explored in an analysis of simulated data ( Supplementary Fig. 4 ). Incorporation of white matter spread of discharges using diffusion tensor imaging is a subject of ongoing work. Nonetheless, the success of our algorithm supports the notion that at least a significant portion of discharge spread is in fact radial. Candidate mechanisms for radial spread include local, isotropic synaptic spread 56-58 and ephaptic spread. 57-60 While our methodology supports the presence of radial spread of discharges, it is unable to elucidate the mechanism of this spread.

Despite the limitations of our approach, our results suggest that both interictal and ictal discharges observed in iEEG recordings represent propagated activity arising from more spatially constrained seizure foci. Seizures may originate from, and also travel to, IED foci. By examining the differences in time of receipt, we are able to localize the hypothesized source of epileptiform activity in both the interictal and ictal data. This approach therefore can have important clinical implications, as it would suggest that localizations can be derived from interictal discharges, which are much more frequent than seizures, and can be used to identify focal sources of activity underlying even the relatively broad epileptiform activity observed in iEEG recordings. More generally, our results provide an approach for better identifying the source and spread of both interictal and ictal activity, an important requirement for better understanding how seizure foci interact with epilepsy networks 61 and for understanding the spatiotemporal evolution of seizures to guide both electrode implantation and targeted surgical interventions. 62

Benjamin Diamond provided intellectual guidance. We are indebted to all patients who have selflessly volunteered their time to participate in this study.

This work was supported by the National Institute for Neurological Disorders and Stroke [ZIA NS003144-09].

J.M.D. reports no disclosures relevant to the manuscript. C.P.W. reports no disclosures relevant to the manuscript. J.I.C. reports no disclosures relevant to the manuscript. S.R. reports no disclosures relevant to the manuscript. S.K.I. reports no disclosures relevant to the manuscript. K.A.Z. reports no disclosures relevant to the manuscript.

Supplementary material is available at Brain online.

Berg AT , Berkovic SF , Brodie MJ , et al.  Revised terminology and concepts for organization of seizures and epilepsies: Report of the ILAE Commission on Classification and Terminology, 2005–2009 . Epilepsia . 2010 ; 51 : 676 – 685 .

Google Scholar

Fisher RS , Boas WVE , Blume W , et al.  Epileptic seizures and epilepsy: Definitions proposed by the International League Against Epilepsy (ILAE) and the International Bureau for Epilepsy (IBE) . Epilepsia . 2005 ; 46 : 470 – 472 .

Wagner FB , Eskandar EN , Cosgrove GR , et al.  Microscale spatiotemporal dynamics during neocortical propagation of human focal seizures . Neuroimage . 2015 ; 122 : 114 – 130 .

Gonzalez-Ramirez LR , Ahmed OJ , Cash SS , Wayne CE , Kramer MA . A biologically constrained, mathematical model of cortical wave propagation preceding seizure termination . PLoS Comput Biol. 2015 ; 11 : e1004065 .

Martinet LE , Fiddyment G , Madsen J , et al.  Human seizures couple across spatial scales through travelling wave dynamics . Nat Commun. 2017 ; 8 : 1 – 13 .

Jy L , Smith EH , Bateman LM , et al.  Multivariate regression methods for estimating velocity of ictal discharges from human microelectrode recordings . J Neural Eng. 2017 ; 14 : 044001 .

Smith EH , Liou JY , Davis TS , et al.  The ictal wavefront is the spatiotemporal source of discharges during spontaneous human seizures . Nat Commun. 2016 ; 7 : 11098 .

Jy L , Smith EH , Bateman LM , et al.  A model for focal seizure onset, propagation, evolution, and progression . Elife . 2020 ; 9 : e50927 .

Diamond JM , Diamond BE , Trotta MS , Dembny K , Inati SK , Zaghloul KA . Travelling waves reveal a dynamic seizure source in human focal epilepsy . Brain . 2021 ; 144 : 1751 – 1763 .

Smith EH , Jy L , Merricks EM , et al.  Human interictal epileptiform discharges are bidirectional traveling waves echoing ictal discharges . Elife . 2022 ; 11 : e73541 .

Schevon CA , Weiss SA , McKhann G Jr , et al.  Evidence of an inhibitory restraint of seizure activity in humans . Nat Commun. 2012 ; 3 : 1060 .

Rossi LF , Wykes RC , Kullmann DM , Carandini M . Focal cortical seizures start as standing waves and propagate respecting homotopic connectivity . Nat Commun. 2017 ; 8 : 1 – 11 .

Proix T , Jirsa VK , Bartolomei F , Guye M , Truccolo W . Predicting the spatiotemporal diversity of seizure propagation and termination in human focal epilepsy . Nat Commun. 2018 ; 9 : 1 – 15 .

de Curtis M , Avanzini G . Interictal spikes in focal epileptogenesis . Prog Neurobiol. 2001 ; 63 : 541 – 567 .

Marsh ED , Peltzer B , Brown MW III , et al.  Interictal EEG spikes identify the region of electrographic seizure onset in some, but not all, pediatric epilepsy patients . Epilepsia . 2010 ; 51 : 592 - 601 .

Sabolek HR , Swiercz WB , Lillis KP , et al.  A candidate mechanism underlying the variance of interictal spike propagation . J Neurosci . 2012 ; 32 : 3009 – 3021 .

Janca R , Krsek P , Jezdik P , et al.  The sub-regional functional organization of neocortical irritative epileptic networks in pediatric epilepsy . Front Neurol. 2018 ; 9 : 184 .

Diamond JM , Chapeton JI , Theodore WH , Inati SK , Zaghloul KA . The seizure onset zone drives state-dependent epileptiform activity in susceptible brain regions . Clin Neurophysiol. 2019 ; 130 : 1628 – 1641 .

Rossi GF . Problems of analysis and interpretation of electrocerebral signals in human epilepsy. a neurosurgeon’s view. In UCLA Forum in medical sciences , volume 17 . Elsevier ; 1973 : 259 – 285 .

Google Preview

Lieb JP , Woods SC , Siccardi A , Crandall PH , Walter DO , Leake B . Quantitative analysis of depth spiking in relation to seizure foci in patients with temporal lobe epilepsy . Electroencephalogr Clin Neurophysiol. 1978 ; 44 : 641 – 663 .

Alarcon G , Guy C , Binnie C , Walker S , Elwes R , Polkey C . Intracerebral propagation of interictal activity in partial epilepsy: Implications for source localisation . J Neurol Neurosurg Psychiatry. 1994 ; 57 : 435 – 449 .

Luders HO , Najm I , Nair D , Widdess-Walsh P , Bingman W . The epileptogenic zone: General principles . Epileptic Disord. 2006 ; 8 ( Suppl. 2 ): S1 – S9 .

de Curtis M , Jefferys JG , Avoli M . Interictal epileptiform discharges in partial epilepsy. Jasper’s basic mechanisms of the epilepsies . 4th Edition. National Center for Biotechnology Information ; 2012 .

Nikitin PV , Martinez R , Ramamurthy S , Leland H , Spiess G , Rao K . Phase based spatial identification of UHF RFID tags. In 2010 IEEE international conference on RFID . IEEE ; 2010 : 102 – 109 .

Mo L , Li C , Xie X . Localization of passive UHF RFID tags on the assembly line. In 2016 International Symposium on Flexible Automation (ISFA) . IEEE ; 2016 : 141 – 144 .

Zhang Y , Amin MG , Kaushik S . Localization and tracking of passive RFID tags based on direction estimation . Int J Antennas Propag. 2007 ; 2007 : 17426 .

Hekimian-Williams C , Grant B , Liu X , Zhang Z , Kumar P . Accurate localization of RFID tags using phase difference. In 2010 IEEE International Conference on RFID . IEEE ; 2010 : 89 – 96 .

Munoz D , Lara FB , Vargas C , Enriquez-Caldera R . Position location techniques and applications . Academic Press ; 2009 .

Milne J . Earthquakes and other earth movements . D. Appleton and Company ; 1886 .

Zhou HW . Rapid three-dimensional hypocentral determination using a master station method . J Geophys Res: Solid Earth . 1994 ; 99 : 15439 – 15455 .

Satriano C , Wu YM , Zollo A , Kanamori H . Earthquake early warning: Concepts, methods and physical grounds . Soil Dyn Earthq Eng . 2011 ; 31 : 106 – 118 .

Matrullo E , De Matteis R , Satriano C , Amoroso O , Zollo A . An improved 1-D seismic velocity model for seismological studies in the Campania–Lucania region (southern Italy) . Geophys J Int. 2013 ; 195 : 460 – 473 .

Saikia U , Rai S . Seismicity pattern, reference velocity model, and earthquake mechanics of south India seismicity pattern, reference velocity model, and earthquake mechanics of south India . Bull Seismol Soc Am 2018 ; 108 : 116 – 129 .

J E Jr , Van Ness P , Rasmussen T , Ojemann L . Outcome with respect to epileptic seizures . Surg Treat Epilepsies . 1993 ; 2 : 609 – 621 .

Brown MW , Porter BE , Dlugos DJ , et al.  Comparison of novel computer detectors and human performance for spike detection in intracranial EEG . Clin Neurophysiol. 2007 ; 118 : 1744 – 1752 .

Gaspard N , Alkawadri R , Farooque P , Goncharova II , Zaveri HP . Automatic detection of prominent interictal spikes in intracranial EEG: Validation of an algorithm and relationship to the seizure onset zone . Clin Neurophysiol. 2014 ; 125 : 1095 – 1103 .

Malinowska U , Badier JM , Gavaret M , Bartolomei F , Chauvel P , Benar CG . Interictal networks in magnetoencephalography . Hum Brain Mapp. 2014 ; 35 : 2789 – 2805 .

Bourien J , Bartolomei F , Bellanger J , Gavaret M , Chauvel P , Wendling F . A method to identify reproducible subsets of co-activated structures during interictal spikes. Application to intracerebral EEG in temporal lobe epilepsy . Clin Neurophysiol. 2005 ; 116 : 443 – 455 .

Tomlinson SB , Bermudez C , Conley C , Brown MW , Porter BE , Marsh ED . Spatiotemporal mapping of interictal spike propagation: A novel methodology applied to pediatric intracranial EEG recordings . Front Neurol. 2016 ; 7 : 229 .

Anderson DN , Charlebois CM , Smith EH , Arain AM , Davis TS , Rolston JD . Probabilistic comparison of gray and white matter coverage between depth and surface intracranial electrodes in epilepsy . Sci Rep. 2021 ; 11 : 1 – 11 .

Hufnagel A , Dumpelmann M , Zentner J , Schijns O , Elger CE . Clinical relevance of quantified intracranial interictal spike activity in presurgical evaluation of epilepsy . Epilepsia . 2000 ; 41 : 467 – 478 .

Asano E , Muzik O , Shah A , et al.  Quantitative interictal subdural EEG analyses in children with neocortical epilepsy . Epilepsia . 2003 ; 44 : 425 – 434 .

Alarcon G , Garcia Seoane J , Binnie C , et al.  Origin and propagation of interictal discharges in the acute electrocorticogram. Implications for pathophysiology and surgical treatment of temporal lobe epilepsy . Brain: J Neurol . 1997 ; 120 : 2259 – 2282 .

Varotto G , Tassi L , Franceschetti S , Spreafico R , Panzica F . Epileptogenic networks of type ii focal cortical dysplasia: A stereo-EEG study . Neuroimage . 2012 ; 61 : 591 – 598 .

Wennberg R , Quesney F , Olivier A , Rasmussen T . Electrocorticography and outcome in frontal lobe epilepsy . Electroencephalogr Clin Neurophysiol. 1998 ; 106 : 357 – 368 .

Barry E , Sussman NM , O’Connor MJ , Harner RN . Presurgical electroencephalographic patterns and outcome from anterior temporal lobectomy . Archives of Neurol . 1992 ; 49 : 21 – 27 .

Verrotti A , Morresi S , Cutarella R , Morgese G , Chiarelli F . Predictive value of EEG monitoring during drug withdrawal in children with cryptogenic partial epilepsy . Neurophys Clin/Clin Neurophysiol . 2000 ; 30 : 240 – 245 .

Fisher RS , Scharfman HE , et al.  How can we identify ictal and interictal abnormal activity? In Issues in clinical epileptology: A view from the bench . Springer ; 2014 : 3 – 23 .

Sammaritano M , Gigli GL , Gotman J . Interictal spiking during wakefulness and sleep and the localization of foci in temporal lobe epilepsy . Neurology . 1991 ; 41 ( 2 Part 1 ): 290 – 290 .

Malow BA , Selwa LM , Ross D , Aldrich MS . Lateralizing value of interictal spikes on overnight sleep-EEG studies in temporal lobe epilepsy . Epilepsia . 1999 ; 40 : 1587 – 1592 .

Emerson RG , Turner CA , Pedley TA , Walczak TS , Forgione M . Propagation patterns of temporal spikes . Electroencephalogr Clin Neurophysiol. 1995 ; 94 : 338 – 348 .

Schlafly ED , Marshall FA , Merricks EM , et al.  Multiple sources of fast traveling waves during human seizures: Resolving a controversy . J Neurosci . 2022 ; 42 : 6966 – 6982 .

Daniel AG , Laffont P , Zhao M , Ma H , Schwartz TH . Optical electrocorticogram (OECoG) using wide-field calcium imaging reveals the divergence of neuronal and glial activity during acute rodent seizures . Epilepsy Behav. 2015 ; 49 : 61 – 65 .

Khoshkhoo S , Vogt D , Sohal VS . Dynamic, cell-type-specific roles for gabaergic interneurons in a mouse model of optogenetically inducible seizures . Neuron . 2017 ; 93 : 291 – 298 .

Zhang X , Qiao Z , Liu N , et al.  Stereotypical patterns of epileptiform calcium signal in hippocampal CA1, CA3, dentate gyrus and entorhinal cortex in freely moving mice . Sci Rep . 2019 ; 9 : 1 – 9 .

Braitenberg V , Schuz A . Cortex: Statistics and geometry of neuronal connectivity . Springer Science & Business Media ; 2013 .

Engel J Jr . Seizures and epilepsy . Volume 83 . Oxford University Press , 2013 .

Codadu NK , Parrish RR , Trevelyan AJ . Region-specific differences and areal interactions underlying transitions in epileptiform activity . J Physiol (Lond). 2019 ; 597 : 2079 – 2096 .

Lian J , Bikson M , Shuai J , Durand DM . Propagation of non-synaptic epileptiform activity across a lesion in rat hippocampal slices . J Physiol (Lond). 2001 ; 537 ( Pt 1 ): 191 .

Zhang M , Ladas TP , Qiu C , Shivacharan RS , Gonzalez-Reyes LE , Durand DM . Propagation of epileptiform activity can be independent of synaptic transmission, gap junctions, or diffusion and is consistent with electrical field transmission . J Neurosci . 2014 ; 34 : 1409 – 1419 .

Spencer SS . Neural networks in human epilepsy: Evidence of and implications for treatment . Epilepsia . 2002 ; 43 : 219 – 227 .

Talairach J , Lesion BJ . “Irritative” zone and epileptogenic focus . Stereotact Funct Neurosurg . 1966 ; 27 ( 1–3 ): 91 – 94 .

  • discharge, body substance
  • epileptiform discharges
  • wave - physical agent

Supplementary data

Email alerts, citing articles via, looking for your next opportunity.

  • Contact the editorial office
  • Guarantors of Brain
  • Recommend to your Library

Affiliations

  • Online ISSN 1460-2156
  • Print ISSN 0006-8950
  • Copyright © 2024 Guarantors of Brain
  • About Oxford Academic
  • Publish journals with us
  • University press partners
  • What we publish
  • New features  
  • Open access
  • Institutional account management
  • Rights and permissions
  • Get help with access
  • Accessibility
  • Advertising
  • Media enquiries
  • Oxford University Press
  • Oxford Languages
  • University of Oxford

Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide

  • Copyright © 2024 Oxford University Press
  • Cookie settings
  • Cookie policy
  • Privacy policy
  • Legal notice

This Feature Is Available To Subscribers Only

Sign In or Create an Account

This PDF is available to Subscribers Only

For full access to this pdf, sign in to an existing account, or purchase an annual subscription.

Advertisement

Can quantum hints in the brain revive a radical consciousness theory?

With anaesthetics and brain organoids, we are finally testing the idea that quantum effects explain consciousness – and the early results suggest this long-derided idea may have been misconstrued

By George Musser

17 January 2024

New Scientist Default Image

TWO weeks before the pandemic lockdown in March 2020, I flew to Tucson, Arizona, and knocked on the door of a suburban ranch-style house. I was there to visit Stuart Hameroff , anaesthesiologist and co-inventor, with Nobel prize-winning physicist Roger Penrose , of a radical proposal for how conscious experience arises: namely, that it has its origins in quantum phenomena in our brains.

Such ideas have existed, in various guises, on the fringes of mainstream consciousness research for decades. They have never come in from the cold because, as their critics argue, there is no solid experimental evidence that quantum effects occur in the brain, never mind a clear idea of how they would give rise to consciousness . “It was very popular to bash us,” Hameroff told me.

Roger Penrose: "Consciousness must be beyond computable physics"

But after a week interrogating the concept with him, I realised that his version of quantum consciousness, at least, is widely misconstrued. Partly, I think that is Hameroff’s fault. He creates the impression of a single take-it-or-leave-it package. In fact, his idea is a series of independent proposals that each force us to confront important questions about the relationship among fundamental physics, biology and that ineffable thing we call consciousness.

Moreover, having seen some experiments that Hameroff was proposing during my visit come to fruition, it has become clear that his ideas can submit to experimental investigation. Researchers have now produced tentative evidence to suggest that fragile quantum states can endure in the brain, and also that anaesthetics have an impact on them.

So is it time to start taking…

Sign up to our weekly newsletter

Receive a weekly dose of discovery in your inbox! We'll also keep you up to date with New Scientist events and special offers.

To continue reading, subscribe today with our introductory offers

No commitment, cancel anytime*

Offer ends 28th October 2023.

*Cancel anytime within 14 days of payment to receive a refund on unserved issues.

Inclusive of applicable taxes (VAT)

Existing subscribers

More from New Scientist

Explore the latest news, articles and features

Learning piano triggers complex changes to your brain's activity

Subscriber-only

The surprising benefits of having an asymmetrical face, body and brain

Psychedelic drug helps treat ptsd and traumatic brain injuries, zapping the brain with electricity makes us easier to hypnotise, popular articles.

Trending New Scientist articles

brain travelling waves

Accessibility links

  • Skip to content
  • Accessibility Help

Follow us on

  • Quick Takes

This is your brain looking at art

The world of innovative tech entrepreneurs and designers.

This is your brain looking at art

How warmer ice creams curb emissions

Clean energy's awkward waste problem

Clean energy's awkward waste problem

Sweden's electric ferry that can fly

Sweden's electric ferry that can fly

Europe's first 'tri-mode' hybrid train

Europe's first 'tri-mode' hybrid train

The portrait covered up for modesty

The portrait covered up for modesty

The luxury cars going electric

The luxury cars going electric

A cheaper way to grow diamonds

A cheaper way to grow diamonds

Inside Cern’s 'antimatter factory'

Inside Cern’s 'antimatter factory'

The future of roads is electric

The future of roads is electric

'Biologically I'm a few hundred ages'

'Biologically I'm a few hundred ages'

Putting a jetpack to the test

Putting a jetpack to the test

When AI has an overactive imagination

When AI has an overactive imagination

The world's richest esports player

The world's richest esports player

How a robot inspired my ballet dancing

How a robot inspired my ballet dancing

How AI could raise IVF success rates

How AI could raise IVF success rates

Your boss may be spying on you

Your boss may be spying on you

Music stored in a painting using DNA

Music stored in a painting using DNA

Virtual reality on the catwalk

Virtual reality on the catwalk

How VR could help investigate fires

How VR could help investigate fires

How to speak a language you can’t

How to speak a language you can’t

Robot teaches itself to ice skate

Robot teaches itself to ice skate

brain travelling waves

Woman anxiously holding man's forearm on an airplane

Here’s what happens to your body during plane turbulence – and how to reduce the discomfort it causes

brain travelling waves

Professor and Director of the Clinical Anatomy Learning Centre, Lancaster University

Disclosure statement

Adam Taylor does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.

Lancaster University provides funding as a founding partner of The Conversation UK.

View all partners

This week has seen another barrage of unsettled weather sweep across the UK, with many flights delayed or cancelled. Some of those who were fortunate enough to take off found themselves arriving at destinations that weren’t on their boarding passes – such as passengers travelling from Stansted to Newquay who eventually diverted to Malaga .

One thing that was consistently described by passengers was that parts of the flights and the attempted landings were some of the most unnerving they’d ever experienced, due to turbulence.

Turbulence results from uneven air movement, which is increasing in frequency. If you turn your hair dryer on at home and hold it still, the air moves at a constant rate, but once you begin drying your hair and moving the hairdryer around, the air movement becomes uneven, that is to say, turbulent.

Although turbulence may be unnerving and make you feel unwell, it is important to recognise that it is very common and typically nothing to worry about if you’re in your seat with your seatbelt fastened.

How the body detects and responds to turbulence

The body recognises itself within any environment. Its relationship with objects in terms of distance and direction is called spatial orientation .

When flying, this is typically moving forwards, ascending, some turns and a descent. However, turbulence disrupts this relationship and confuses the sensory information being received by the brain – it makes the body want to respond or recalibrate.

A plane flying under dark clouds.

Our inner ears play a pivotal role in all this. It consists of complex apparatuses that undertake more than hearing. These include the cochlea, three semi-circular canals , the utricle and the saccule .

The cochlea is responsible for hearing. It converts sound energy into electrical energy that is then “heard” by the brain. The remaining structures are responsible for the balance and position of the head and body. The semi-circular canals are positioned in a vertical (side to side), horizontal and front-to-back plane, detecting movement in a nodding, shaking and touching ear-to-shoulder direction.

Attached to these canals are the utricle and saccule , which can detect movement and acceleration .

All of these apparatuses use microscopic hair cells in a specialised fluid called endolymph that flows with the head to create a sense of movement. When the plane encounters turbulence, this fluid moves around, but unpredictably. It takes about ten to 20 seconds for the fluid to recalibrate its position, while the brain struggles to understand what is going on.

When the aircraft hits turbulence, the balance apparatus cannot distinguish the movement of the plane from that of the head, so the brain interprets the aircraft movement as that of the head or body. But this doesn’t match the visual information being received, which causes sensory confusion.

The reason the inner ear causes so much confusion is because during flights you are devoid of your primary sensory tool relative to the external environment – your sight and the horizon.

Eighty per cent of spatial information comes from your eyes during flight. However, you only have the seat in front of you or the cabin as a reference point, which means your inner ear becomes the dominant sensory message to the brain during turbulence and disrupts the “vestibulo-ocular reflex” . This reflex keeps your vision aligned with your balance or expected position.

Vision is the most valued of the senses and one-third of the brain is attributed to its function, reinforcing its importance in spatial orientation.

This sensory mixed messaging often results in things like dizziness and sweating as well as gastrointestinal symptoms, such as nausea and vomiting .

Motion sickness can be triggered by turbulence and although research into specific airsickness is limited, other modes that induce motion sickness suggest that women are more susceptible than men, particularly in the early stages of the menstrual cycle.

The turbulence also causes an increase in your heart rate, which is already higher than normal when flying because of a decrease in oxygen saturation .

What about the pilots?

Commercial pilots accrue thousands of hours at the controls, they are subject to the same forces as the passengers.

Over time, they can adapt to these forces and experiences , but they also have a couple of additional resources that most passengers don’t.

They have the view out of the cockpit windows, so have a horizon to use as a reference point and can see what lies immediately ahead.

If it is cloudy or visibility is low, their instruments provide additional visual reference to the position of the aircraft. This doesn’t mean they are immune to the effects of turbulence, with some studies reporting up to 71% of trainee pilots reporting episodes of airsickness.

How to reduce the discomfort

A window seat can help, or even looking out the window. This gives the brain some sensory information through visual pathways, helping calm the brain in response to the vestibular information it is receiving.

If you can get one, a seat towards the front or over the wing reduces the effects of turbulence.

Deep or rhythmical breathing can help reduce motion sickness induced by turbulence. Focusing on your breathing calms the nervous system .

Don’t reach for the alcohol. While you may feel it calms your nerves, if you hit turbulence it’s going to interfere with your visual and auditory processing and increase the likelihood of vomiting.

If you suffer from motion sickness and are worried about turbulence while flying, then there are also drugs that can help , including certain antihistamines .

Finally, it’s important to remember that although turbulence can be unpleasant, aircraft are designed to withstand the forces it generates and many passengers, even frequent fliers, will rarely encounter the most severe categories of turbulence because pilots actively plan routes to avoid it.

brain travelling waves

Business Ethics – Assistant/Associate Professor

brain travelling waves

Professor/Associate Professor in Environmental Informatics

brain travelling waves

Project Manager

brain travelling waves

Senior Lecturer/ Associate Professor/ Professor in Social Work

brain travelling waves

Director, Strategic Events

IEEE Pulse

Treating the Brain With Focused Ultrasound

  • January 23, 2024

Clinicians have long used ultrasound to image inside the body, but it may prove even more useful as a therapeutic tool

Focused ultrasound is an early stage, noninvasive therapy with the potential to treat a range of medical conditions. Like diagnostic ultrasound, it uses sound waves above the range of human hearing. But its purpose is to interact with tissues in the body, rather than just produce images of them. In focused ultrasound, multiple, intersecting beams of high frequency sound are aimed to converge on specific targets deep within the body. There, the ultrasound energy can act in multiple ways to either modify or destroy tissue.

The U.S. Food and Drug Administration (FDA) has approved focused ultrasound to treat eight conditions, including uterine fibroids, cancer that has spread into the bones, and breast, prostate, and liver tumors. It is under investigation for dozens of other applications [1]. However, it is the potential of focused ultrasound to treat diseases of the brain that is generating the most excitement. Already approved for the treatment of Parkinson’s disease and essential tremor, researchers are exploring focused ultrasound for a variety of other neurological and psychiatric conditions.

“The brain has taken over focused ultrasound,” says Elisa Konofagou, professor and principal investigator at the Ultrasound and Elasticity Imaging Laboratory at Columbia University (Figure 1). “Neurosurgeons are embracing it, as well as neurologists, oncologists, and psychiatrists. There are many devastating brain diseases that could potentially benefit from ultrasound.”

Treating the Brain With Focused Ultrasound

Figure 1. Elisa Konofagou is a professor in BME and a principal investigator at the Ultrasound and Elasticity Imaging Laboratory, Columbia University. (Photo courtesy of Konofagou.)

Heat as a tool

Focused ultrasound can be made to produce different therapeutic effects by varying the parameters of the ultrasound beams. In high intensity focused ultrasound (HIFU), high power sound waves converge at one point to raise the temperature and destroy tissue. Typically, magnetic resonance imaging is used to identify and target the specific tissue to be treated and guide the treatment in real time. This technique is FDA-approved and in clinical use for several applications.

The ability of HIFU to precisely ablate tissue without requiring surgery makes it a promising alternative to deep brain stimulation (DBS) for movement disorders associated with Parkinson’s disease and essential tremor. Although DBS of targets in the thalamus can be effective for these conditions, it is a costly and invasive treatment.

“We know that the cells in that part of the brain fire abnormally and are entrained to fire with the tremor,” says Nir Lipsman, a neurosurgeon at Sunnybrook Research Institute in Canada (Figure 2). “Getting rid of that neuronal activity, whether with electricity or heat, is a way to impact that tremor. High intensity focused ultrasound is a new tool that we use in an established way—to generate a lesion in the brain.”

Treating the Brain With Focused Ultrasound

Figure 2. Nir Lipsman is a neurosurgeon at the Sunnybrook Research Institute, Canada. (Photo courtesy of Sunnybrook Health Sciences Centre.)

Over the last ten years, Lipsman has helped develop several clinical trials of focused ultrasound. He was part of the team that established HIFU as an effective and safe treatment for patients with essential tremor, leading to FDA-approval in 2016 [2]. A recent follow-up study demonstrated that patients who underwent HIFU for essential tremor sustained improvements for up to five years [3].

Currently, HIFU is FDA-approved and commercially available for the treatment of essential tremor and tremor-dominant Parkinson’s disease. It is also under investigation for other conditions where overactivity in a specific brain region causes problems, such as chronic pain and obsessive-compulsive disorder [4], [5].

Dialing down the intensity

In contrast to HIFU that destroys cells, focused ultrasound at low intensities, known as low intensity focused ultrasound (LIFU), is under investigation for its ability to modulate neuronal activity without incurring cell damage. There is mounting interest in the use of LIFU to treat neurological and psychiatric disorders, but this technique is still experimental [4], [5].

There are clear targets for some conditions, says Noah Philip, a professor at Brown University and the founding section chief of psychiatric neuromodulation at VA Providence (Figure 3). For instance, he is conducting trials of LIFU to inhibit brain areas known to be involved in post-traumatic stress disorder, anxiety, and depression. “If we can noninvasively reach deep into the brain and turn off these areas that drive psychiatric illnesses, the hope is that it will yield a treatment that is more effective than what we have right now,” says Philip.

Treating the Brain With Focused Ultrasound

Figure 3. Noah Philip, professor at Brown University and the founding section chief of psychiatric neuromodulation at VA Providence. (Photos courtesy of Brown University.)

That is also the hope of Wynn Legon, an assistant professor and a principal investigator at the Fralin Biomedical Research Institute at the Virginia Tech Carilion School of Medicine (Figure 4). Legon has been at the forefront of LIFU research for a decade and his lab is currently focusing on investigating applications for addiction and chronic pain. “We’ve done a lot of studies in healthy control participants demonstrating its safety and tolerability and showing that it can reduce their perceived pain,” he says. “Now we want to see how durable the effect is and if it will impact quality of life or reduce pain for patients with chronic pain. We are just on the cusp of getting into some clinical indications.”

Treating the Brain With Focused Ultrasound

Figure 4. Wynn Legon is an assistant professor and a principal investigator at the Fralin Biomedical Research Institute at the Virginia Tech Carilion School of Medicine. (Photo courtesy of Clayton Metz for Virginia Tech.)

Shaking open barriers

In addition to manipulating cells deep in the brain, LIFU shows promise in another application: opening the blood–brain barrier. This protective membrane surrounds a specialized system of blood vessels in the brain, preventing pathogens and other dangerous substances in the blood from gaining entrance. However, it can also prevent drugs and other therapies from reaching the brain, complicating the treatment of brain tumors and Alzheimer’s disease, among other conditions.

Researchers are now leveraging the fact that LIFU mechanically stimulates anything under its focus to disrupt the blood–brain barrier. This is achieved by injecting microbubbles into the bloodstream and then applying LIFU to them as they pass through blood vessels in the brain. The ultrasound activates the bubbles, temporarily loosening up the blood–brain barrier, providing an opportunity for drugs or other therapeutic molecules to pass through.

Konofagou’s group recently demonstrated the use of LIFU to open the blood–brain barrier and enable genome editing [6]. Gene therapy is a sought-after goal for neurologic diseases, such as Alzheimer’s and Parkinson’s, but getting gene editing vectors across the blood–brain barrier has proven challenging. Konofagou used LIFU and microbubbles to open the blood–brain barrier and deliver CRISPR-encoded viral vectors into the brains of mice. The technique increased gene editing efficiency in mouse neurons, suggesting that it could one day be used to edit the genome of neuronal cells to correct genes involved in brain diseases.

Konofagou’s research has also shown that opening the blood–brain barrier with ultrasound may have therapeutic effects on its own. “Opening the blood–brain barrier stimulates immune cells in the brain called microglia that are mechanically sensitive,” says Konofagou. “They respond to the ultrasound by clearing away more debris in the brain, such as the proteins beta-amyloid and tau that accumulate in Alzheimer’s disease.” In a recent experiment, mice treated with LIFU alone had decreased levels of beta-amyloid and tau and improved working memory. When the researchers applied the technique to individuals with Alzheimer’s in a clinical trial, they found a modest reduction in beta-amyloid in the region where the blood–brain barrier was opened compared to an untreated area [7].

Challenges and unknowns

The potential of focused ultrasound to provide noninvasive access to the brain and treat previously intractable diseases has produced a lot of buzz, but there are limitations to the technology. Currently, HIFU’s usefulness is limited to making lesions in areas near the center of the brain. In addition, not everyone’s skull is amenable to HIFU; denser skulls can absorb much of the ultrasound energy and make it difficult to create lesions in some patients, says Lipsman. “These are some of the technical limitations to the technology that you learn about once you start using a tool widely,” he says. “But we and others are working on overcoming these limitations to make HIFU as widely accessible as possible.”

When it comes to LIFU, basic questions remain about how the treatment produces its modulatory effects on cells. Legon suspects mechanical forces are important. “There is evidence showing that some channels in brain cells that have been traditionally regarded as being voltage-gated electrical channels also respond to mechanical energy,” he says. “The idea is that if you send in a pressure wave across a cell membrane or to one of these specific channels, it will perturb the channel mechanically and that can bias it to either open or close.”

While the exact mechanisms underlying how LIFU works are under active investigation, our shifting understanding of how psychiatric illnesses work presents another challenge, says Philip. Today, psychiatric illnesses are thought of as disruptions of large-scale, functional neural networks, marked by unhealthy patterns of communication between different brain areas. “We have a highly precise and specific intervention device and yet, we are becoming more mindful that these illnesses are a product of large-scale neural networks,” says Philip. “We hope that modulating core elements of these networks is going to be helpful, but it is entirely possible that LIFU may not be sufficient because it is, in fact, too precise a tool.”

Bright future

Despite the current unknowns, Philip and others are optimistic about the future of focused ultrasound. “As a psychiatrist as well as a researcher, I’m acutely mindful that the treatments we have today simply do not meet our patients’ needs,” he says. “This technology, unlike some other neuromodulatory interventions such as transcranial magnetic stimulation or deep brain stimulation, has the potential to be extremely portable and scalable.”

According to the Focused Ultrasound Foundation, there are nearly 40 neurological and psychiatric applications currently under investigation; most are in early stages. In addition to the brain disorders already mentioned, this includes depression, epilepsy, schizophrenia, bipolar disorder, and amyotrophic lateral sclerosis [1]. Much work remains to establish where and how the technology will provide the most therapeutic value. “The great advantage of focused ultrasound is the ability to reach these previously inaccessible areas of the brain noninvasively,” says Legon. “I think it has promise for any clinical indication where there is an established location deep in the brain.”

In addition, Lipsman looks forward to future studies to investigate how best to take advantage of the transient blood–brain barrier opening that LIFU induces. “We are working closely with our partners in the drug development world to determine which drugs we should deliver to the brain for different neurological diseases,” he says. “Whether it’s gene therapy, or anti-amyloid or anti-tau antibodies for Alzheimer’s, or enzymes for Parkinson’s—this pairing of pharmaceutical or therapeutic development with the technology is an exciting area.”

“Although we are very much in the early days of using focused ultrasound to treat brain diseases, I think there is a bright future for this technology,” Lipsman adds. “We are learning a lot about the hardware and software, as well as the optimal indications for focused ultrasound.” Given the challenges associated with accessing the brain and the limitations of some current treatments—as well as the mounting efforts to optimize the technology—the excitement surrounding focused ultrasound’s potential is warranted.

  • Focused Ultrasound Foundation: Diseases and Conditions . Accessed: Oct. 15, 2023. [Online]. Available: https://www.fusfoundation.org/diseases-and-conditions/
  • W. J. Elias et al., “A randomized trial of focused ultrasound thalamotomy for essential tremor,” New England J. Med. , vol. 375, no. 8, pp. 730–739, Aug. 2016, doi: 10.1056/NEJMoa1600159.
  • G. R. Cosgrove et al., “Magnetic resonance imaging-guided focused ultrasound thalamotomy for essential tremor: 5-year follow-up results,” J. Neurosurgery , vol. 138, no. 4, pp. 1028–1033, Aug. 2022, doi: 10.3171/2022.6.JNS212483.
  • Y. Meng, K. Hynynen, and N. Lipsman, “Applications of focused ultrasound in the brain: From thermoablation to drug delivery,” Nature Rev. Neurol. , vol. 17, no. 1, pp. 7–22, Jan. 2021, 10.1038/s41582-020-00418-z.
  • H. Baek et al., “Clinical intervention using focused ultrasound (FUS) stimulation of the brain in diverse neurological disorders,” Frontiers Neurol. , vol. 13, May 2022, Art. no. 880814. doi: 10.3389/fneur.2022.880814.
  • Y.-H. Lao et al., “Focused ultrasound-mediated brain genome editing,” Proc. Nat. Acad. Sci. USA , vol. 120, no. 34, Aug. 2023, Art. no. e2302910120, doi: 10.1073/pnas.2302910120.
  • M. E. Karakatsani et al., “Focused ultrasound mitigates pathology and improves spatial memory in Alzheimer’s mice and patients,” Theranostics , vol. 13, no. 12, pp. 4102–4120, Jul. 2023, doi: 10.7150/thno.79898.

For IEEE Members

Ieee spectrum, follow ieee spectrum, support ieee spectrum, enjoy more free content and benefits by creating an account, saving articles to read later requires an ieee spectrum account, the institute content is only available for members, downloading full pdf issues is exclusive for ieee members, access to spectrum 's digital edition is exclusive for ieee members, following topics is a feature exclusive for ieee members, adding your response to an article requires an ieee spectrum account, create an account to access more content and features on ieee spectrum , including the ability to save articles to read later, download spectrum collections, and participate in conversations with readers and editors. for more exclusive content and features, consider joining ieee ., join the world’s largest professional organization devoted to engineering and applied sciences and get access to all of spectrum’s articles, archives, pdf downloads, and other benefits. learn more →, access thousands of articles — completely free, create an account and get exclusive content and features: save articles, download collections, and talk to tech insiders — all free for full access and benefits, join ieee as a paying member., this engineer’s hardware is inspired by the brain, manan suri’s neuromorphic systems run ai on sensors, drones, and vr headsets.

A man in a suit standing in front of a large machine that has yellow wires coming from it.

Manan Suri displays a wafer-level testing system for unpackaged chips and devices built by researchers at the Indian Institute of Technology Delhi.

In work and in life, it’s easy to get stuck in your ways. That’s why Manan Suri has always looked to expand his horizons both professionally and personally.

Growing up in India, he was used to transitions and new experiences because his family frequently moved around the country as his father relocated for his job as a chemical engineer. Traveling stuck with Suri in adulthood. He studied and worked in Dubai, the United States, France, and Belgium over the course of his twenties.

Indian Institute of Technology Delhi

Occupation:

Associate professor and founder of Cyran AI Solutions, New Delhi

Bachelor’s and master’s degrees in electrical and computer engineering, both from Cornell; Ph.D. in nanoelectronics from the CEA-Leti research institute in Grenoble, France

Eventually, Suri moved back to India to become an assistant professor at the Indian Institute of Technology Delhi . There he set up a research group focused on developing brain-inspired (neuromorphic) computer hardware for low-power devices like sensors, drones, and virtual-reality headsets. He is now an associate professor.

He also launched a startup to commercialize his lab’s expertise: Cyran AI Solutions , based in New Delhi, works with companies and government agencies on a variety of projects. These include automating the inspection process for identifying defects in semiconductors and developing computer-vision systems to improve crop yields and analyze geospatial Earth-observation data.

While balancing a career in academia and industry is challenging, Suri says, he relishes the opportunity to constantly learn.

“Once I’ve figured out how a system works, I start getting bored,” he says.

Suri, an IEEE member, believes that embracing change is a key ingredient for success. This is what has driven him to continually move on to new projects, push into new disciplines, and even move from country to country to experience a different way of life.

“It accelerates your ability to learn new things,” he says. “It puts you on a fast trajectory and helps shed some of your inhibitions or get over the inertia in what you’re doing or how you’re living.”

Inspired by Cornell’s semiconductor lab

Growing up, Suri’s passion was physics, but he quickly realized he was drawn more to the practical applications than theory. This led to a fascination with electronics.

In 2005 he initially enrolled at the Birla Institute of Technology and Science , Pilani, in India, and studied electronics and instrumentation at the institute’s campus in Dubai. After his second year, he transferred to Cornell , in Ithaca, N.Y. His first six months living in the United States, acclimating to a new culture and a different academic environment, were overwhelming, Suri says. What hooked him were Cornell’s high-end facilities available to students studying semiconductor engineering and nanofabrication—in particular, the industry-grade semiconductor clean rooms.

He earned a bachelor’s degree in electrical and computer engineering in 2009 and a master’s degree in the same subject the following year.

New skills in computational neuroscience

After graduating, Suri received offers for Ph.D. positions in the United States and Europe to work on conventional electronics projects. But he didn’t want to get pigeonholed as a traditional semiconductor engineer. He was intrigued by an offer to study neuromorphic systems at the CEA-Leti research institute in Grenoble, France. He was also eager to broaden his life experience and get a taste of the European way of doing things.

The work would push Suri to develop new skills in computational neuroscience and computer science. In 2010 he started a Ph.D. program in the institute’s Advanced Memory Technology Group. There he worked on low-power AI hardware that uses new kinds of nonvolatile memory to emulate how biological synapses process data. This involved using phase-change memory and conductive-bridging RAM to create neural networks for visual pattern extraction and auditory pattern sensitivity .

Suri discovered that his experience with electronics allowed him to approach neuromorphic engineering problems from an entirely different angle than his colleagues had considered. Experts can develop fairly rigid and conventional ways of thinking about their own field, he says, but when those with different skill sets apply them to the same problems, it can often lead to more innovative thinking.

“You bring a completely different perspective,” he says. “It leads to a lot of creativity.”

Setting up his own research lab

After finishing his doctorate in nanoelectronics, Suri got a job working on high-voltage transistors for automotive applications at the semiconductor designer NXP Semiconductors , in Brussels. Since his role was to take a project all the way from concept to fabrication, it was as close to pure research as he could get in industry. But as interesting as the work was, Suri says, he missed the intellectual freedom of academia.

When the opportunity of setting up his own lab at IIT Delhi came along, he jumped at it. He had also been away from his home country for almost a decade and wanted to be closer to family and contribute to the Indian science and technology ecosystem, he says.

“Most users don’t really care about what technology we are using. They just want functional performance at the most cost-effective price.”

“Moving abroad was more a matter of collecting experiences and seeing how different places work,” he says.

Suri’s group at IIT Delhi has made contributions to AI hardware, neuromorphic hardware, and hardware security. The group collaborates with industry research teams around the world, including Meta Reality Labs , Tata Consultancy Services , and GlobalFoundries .

Launching a startup

Despite returning to academia, Suri says he has always been interested in developing practical solutions to real-world challenges, and this goal has guided his research. Whatever project he works on, he always asks himself two questions: Will it solve a real problem? And will someone buy it?

Suri launched his startup in 2018 to turn some of his lab’s work in AI and neuromorphic hardware into commercial products. Cyran AI Solutions’ customers hire the company to solve a range of problems. These have included computer-vision systems for detecting defects in computer chips; hyperspectral data-analysis algorithms designed to run in real time on chips for crop-inspection drones; and AI systems for small, low-power devices and challenging environments like satellites.

While Cyran makes use of its neuromorphic expertise for some problems, it often uses more mature and simpler-to-deploy machine-learning approaches.

“Most users don’t really care about what technology we are using,” Suri says. “They just want functional performance at the most cost-effective price.”

One of the biggest lessons Suri learned from running a startup is to consider the market being served. For earlier projects, he says, the company often devised a solution that was specific to just one customer’s needs and couldn’t be repurposed for other uses. To create a sustainable business, he realized he needed to develop generic solutions that could be deployed more broadly.

“Running Cyran has been like [pursuing a] mini-MBA,” he says. “You need to really pay attention to the market aspects and not just the technology.”

In 2018, MIT Technology Review named Suri one of its 35 Innovators Under 35 for his work on neuromorphic computing.

The need to be hands-on

Keeping a foot in both academia and industry can be challenging, Suri says. Facing resource crunches, whether in time, staffing, or funding, is common. The only way he’s able to manage things is to plan extensively and remain nimble, building in contingencies.

If you can manage it, Suri says, having your fingers in many pies can have major benefits. In particular, working on problems that bridge several disciplines can help you break out of rigid thinking and come up with novel solutions.

It’s not possible to dedicate equal amounts of time to learning every area, he says, so he advises up-and-coming engineers to carefully pick the topics that are most likely to advance their progress. It’s also crucial to dive in and get your hands dirty, rather than focusing on theory, initially.

“Take the plunge and try and figure it out,” he recommends. “As the problem unravels, then you can start getting into the theory or the more formal aspects of the project. You also start to appreciate learning more about the theory as it gets more hands-on.”

  • 35 Years Ago, Researchers Used Brain Waves to Control a Robot ›
  • This Engineer Is Helping to Make India a Global Semiconductor Hub ›
  • A neuron model for efficient AI systems- The New Indian Express ›
  • Manan Suri | MIT Technology Review ›

Edd Gent is a freelance science and technology writer based in Bengaluru, India. His writing focuses on emerging technologies across computing, engineering, energy and bioscience. He's on Twitter at @EddytheGent  and email at edd dot gent at outlook dot com. His PGP fingerprint is ABB8 6BB3 3E69 C4A7 EC91 611B 5C12 193D 5DFC C01B. His public key is here. DM for Signal info.

Why Rip Off Creatives, If Generative AI Can Play Fair?

Satellite tech opens new window on antarctic, video friday: medusai, related stories, from cleaning offices to designing ev charging stations, imagineer morgan pope uses electromagnetism to spark emotions, from hockey enforcer to high-flying evtol ceo.

GlobalData

Insect-inspired AI: how “natural intelligence” could shape the third wave

M achine learning models seek to mimic the natural, human ability to process information, enabling technology to take decisions and complete tasks accordingly – and making artificial intelligence (AI) titularly accurate. However, software company Opteran is pursuing a new approach: one of ‘natural intelligence’ modelled on the neural networks of insects.

Artificial Intelligence versus 'natural intelligence'

Since OpenAI released ChatGPT in 2022, setting the record for the fastest sign up of one million users and supercharging the AI hype, the sector has seen rapid growth and investment. GlobalData predicts that the AI market will be worth $383.3bn by 2030, which would represent a 21% CAGR from 2022.

Explaining the limitations AI faces and the potential solutions offered by a ‘natural intelligence’ model, James Marshall, co-founder and chief science officer at Opteran, tells Verdict: “The difference between a natural intelligence algorithm and an AI algorithm - a deep net - would be that deep nets are usually like a big soup. It's very loose brain inspiration that is inspired by a very small part of your brain - the visual cortex.

“Having found that that simple model of that part of the brain can do image recognition, the industry has then tried to turn that into a silver bullet that solves all the problems in autonomy, and they've had quite a lot of success. However, the challenge is really that most of the success is actually driven by having more data and more compute available. The inputs that are going in are pretty much exponential in growth, whereas the performance improvements are normally linear; that's why people say AI is running out of steam, because you're reaching the limits of how big you can make a network or how much data you can push through it.

“Natural intelligence is exactly trying to solve that more holistic problem … It's moving from having a soup of very simple neurons, which is how you can think about machine learning working, to something much more structured and hence much more efficient with a lot of variety in connection patterns, in the types of neurons and jobs they're doing. It's just trying to capture that real variety we see in real brains, which must be there for a reason.

“If you could solve the autonomy problem with just a soupy brain full of all the same neurons just wired together randomly, then that's how brains would look, but they don't work that way.”

Honeybee foragers and the ‘third wave’ of AI

AI progressed with the development of machine learning in the 1980s, when the approach moved away from logic programming (the ‘first wave’ of AI), and towards the neural network model that enables computers to learn independently. This ‘second wave’ is described by Marshall as “a kind of cartoon of how the brain works”, analysing data and compute to reach decisions.

Opteran posits that insects could provide an alternative, more efficient – and ultimately more intelligent – model to the “soup of very simple neurons”. Marshall suggests that this approach will shape the ‘third wave,’ translating the structure and function of brains into technology.

“People call it third-wave AI, but we call it natural intelligence, to differentiate it from artificial intelligence, which is mainly deep learning, in most people's minds,” he says.

A University of Sheffield spin-out , Opteran studied the brains of various insects with a team of computational and wet neuroscientists, behavioural biologists and computer scientists. Marshall explains: “We spent a lot of time looking at insect brains - which are tiny, they're like a million neurons for a honeybee forager … it can fly about 10km out from the nest in the complex, visual world that we all live in, find something of interest, but then crucially get back to where it came from, and then even communicate where it has been to others – other nest mates – who can then fly out to exactly the same flower patch. They're amazing visual navigators, and they're doing that with a brain of a million neurons that occupies a cubic millimetre.

“Now, if you compare that to where our technology currently is, in terms of mapping, localization and autonomy in general, it's just incomparable. We don't have any technology that can reach that level of performance. So, the question I started off trying to answer within the university – and what led to the formation of Opteran – was: how can a bee brain do that?

“It's not just bees, other insects do it – desert ants, for example. They live in the desert where they can't lay pheromones, which is what people often think about ants doing to navigate around, because the pheromones evaporate as it's too hot there, so they're amazing visual navigators as well … We started off trying to unravel this and we use tools like brain atlases that research groups around the world have worked on, neural recordings (which we did ourselves), we even put animals in VR – we put bees in virtual reality, for example. We kind of pieced together quite a lot of the puzzle around how that tiny brain can actually solve this visual navigation problem.”

'Natural intelligence' solving problems in practice

Current AI-powered robots still require some human direction to complete a task and are at risk of traffic jams or collisions unless software advancements can enable them to make local decisions. Marshall believes Opteran’s model of ‘natural intelligence’ offers a potential avenue for AI to develop in this way.

He explains: “Brains evolved to solve movement, and everything else that we do that's more intelligent than movement sits on top of that basic functionality. We're starting with the same problem; we've looked at the lower-level brain functions that govern navigation. That's good because, as well as being the starting point for autonomy, it's also a hugely challenging technological area with massive commercial demand.”

The technology has potential in a variety of sectors, but the first commercial uses look likely to be in warehouse robotics, which is proving to be a rapidly growing sector with huge demand. According to GlobalData analytics, sales of industrial robots reached $20.7bn in 2022, making it equivalent to 33% of the $63bn market. By 2030, GlobalData predicts that the industrial robots sector will be worth $45.1bn.

 “We're looking at, for example, right now, warehouse robots, drones …” says Marshall, considering the future of  Opteran. “Really, we're selling the mind, which is a set of algorithms, and that mind can control a drone as easily as it can control a legged robot like a Boston Dynamics dog, or a wheeled robot, like a robot vacuum cleaner, or a driverless car.

“We think of it as a ubiquitous technology, but we're commercial, so we'll go after the biggest, easiest commercial prizes early on, and there's an awful lot around warehouse robotics, for example, because there's huge unsatisfied demand there. The technology is really expensive to deploy, and not robust enough … The robustness and the efficiency of your system is absolutely essential at the cost of deploying.”

"Insect-inspired AI: how “natural intelligence” could shape the third wave" was originally created and published by Verdict , a GlobalData owned brand.

The information on this site has been included in good faith for general informational purposes only. It is not intended to amount to advice on which you should rely, and we give no representation, warranty or guarantee, whether express or implied as to its accuracy or completeness. You must obtain professional or specialist advice before taking, or refraining from, any action on the basis of the content on our site.

Honeybee collecting pollen.

  • Skip to main content
  • Keyboard shortcuts for audio player

Short Wave

  • LISTEN & FOLLOW
  • Apple Podcasts
  • Google Podcasts
  • Amazon Music
  • Amazon Alexa

Your support helps make our show possible and unlocks access to our sponsor-free feed.

Experiencing racism may physically change your brain

Jon Hamilton 2010

Jon Hamilton

Rachel Carlson

Rebecca Ramirez, photographed for NPR, 6 June 2022, in Washington DC. Photo by Farrah Skeiky for NPR.

Rebecca Ramirez

Female researcher holds up a model of a human brain.

Scientists know that Black people are at a greater risk for health problems like heart disease , diabetes and Alzheimer's disease than white people. A growing body of research shows that racism in health care and in daily life contributes to these long-standing health disparities for Black communities.

Now, some researchers are asking whether part of the explanation involves how racism, across individual interactions and systems, may physically alter the brain.

How poverty and racism 'weather' the body, accelerating aging and disease

Shots - Health News

How poverty and racism 'weather' the body, accelerating aging and disease.

"That could be behaviors like, let's say, a woman clutching her purse as a black man is walking next to her. Or they could be verbal, like someone saying, like... 'I didn't expect you to be so articulate,'" says Negar Fani , a clinical neuroscientist at Emory University who studies people experiencing Posttraumatic Stress Disorder, or PTSD.

Recently, Fani has collaborated with Nate Harnett , an assistant professor of psychiatry at Harvard Medical School, to study how the brain responds to traumatic events and extreme stress, including the events and stress related to racism.

Individual insights to systemic issues

So how does one go about measuring the impact of zoomed out, societal-scale issues on the individual?

Harnett is the first to admit, it's not the simplest task.

"It's very difficult for neuroimaging to look specifically at redlining," notes Harnett.

But he can—indirectly.

For example, Harnett has used inequities in neighborhood resources as a way of tracking or measuring structural racism.

"We're able to look at these sort of proxy measures in these outcomes of structural racism and then correlate those with both brain and behavioral responses to stress or trauma and see how they tie with different psychiatric disorders like PTSD," Harnett says.

In other research, Harnett and Fani have looked at correlations between racial discrimination and the response to threat in Black women who had experienced trauma. Fani says patients who experience PTSD tend to be more vigilant or show hyperarousal and be startled easily. Fani says their bodies are in a constant state of fight or flight—even when they're in a safe situation.

Racism Is Literally Bad For Your Health

You, Me And Them: Experiencing Discrimination In America

Racism is literally bad for your health.

But in patients who've also experienced racial discrimination, Fani says she sees the opposite effect: They show an increased activation in areas related to emotion regulation.

In some ways, Fani says this activation can be adaptive. For example, people may experience microaggressions or discrimination at work and need to regulate their emotional response in order to get through the moment. But when people have to utilize this strategy over long periods of time, Fani and Harnett think it may contribute to the degradation they've seen in other areas in the brain .

"There's no such thing as a free lunch when it comes to the brain," Harnett says. "Energy has to come from somewhere. And what we think ends up happening is, you know, energy that's reserved for other processes then gets taken away."

Harnett says some people have called this process "weathering," where the stress related to repeated exposure to traumatic experiences erodes parts of the brain.

"Over time, that might contribute to other downstream health problems like cardiovascular disease or diabetes," he says.

Making The Case That Discrimination Is Bad For Your Health

Code Switch

Making the case that discrimination is bad for your health, societal challenges reflected in publishing.

Fani and Harnett say they've faced challenges in publishing their research that seem to reflect the resistance in the medical field to acknowledging health disparities for minority communities.

Fani says that as part of her research, she's administered trauma inventories, which ask patients to recount instances of physical or emotional harm, for decades. When she's used these inventories in papers, they "were never criticized for what they were." But when attempting to publish papers using racial discrimination inventories, she says she's noticed skepticism and a double standard.

"There is a different standard that we're held to," she says. "Like, 'How do you know that people really experienced racism? How do you know that these aren't people who are just more sensitive to slights?'"

While she and Harnett both say they've seen improvements in their field, they hope their work will push other institutions toward a better understanding of how racism can change the brain.

Scientists Reach Out To Minority Communities To Diversify Alzheimer's Studies

Scientists Reach Out To Minority Communities To Diversify Alzheimer's Studies

"Many times in the medical community, you know, racism hasn't been recognized as the kind of insidiously damaging stressor that it is," Fani says. "The hope is that this kind of brain research can lend a kind of legitimacy to racism as a potent social stressor that has a very clear and pronounced effect on the brain."

Listen to Short Wave on Spotify , Apple Podcasts and Google Podcasts .

Today's episode was produced by Rachel Carlson. It was edited by Rebecca Ramirez. Rebecca also fact-checked alongside Rachel. Maggie Luthar was the audio engineer.

  • health disparities
  • racial discrimination

Mountain View, CA

brain travelling waves

Mountain View

Around the Globe

Hurricane tracker, severe weather, radar & maps, news & features, winter center, news / severe weather.

'Rogue' or 'sneaker?' What caused the giant wave in the Marshall Islands

By Jesse Ferrell , AccuWeather meteorologist and senior weather editor

Published Jan 25, 2024 11:29 AM PST | Updated Jan 25, 2024 11:38 AM PST

A series of large waves fueled by an offshore storm flooded the northern portions of the island of Roi-Namur, Kwajalein Atoll, Marshall Islands, damaging the Army base and Freeflight International Airport.

You've probably seen the viral video of a large wave inundating a building. A triple-screen view from several cellphone cameras proves even more impressive. The big wave happened on Saturday, Jan. 20, 2024, on the island of Roi-Namur, part of the Kwajalein Atoll, in the Marshall Islands. The wave caused significant damage to Dyess Army Field and Freeflight International Airport.

Kwajalein Atoll -- colloquially referred to as "Kwaj" by residents -- is a ring of islands in the Pacific Ocean, roughly 1,500 miles northeast of Papua New Guinea. The atoll contains only 3.6 square miles of land -- less than one-fifth the size of Bermuda. Eleven of the 97 islands, including Kwajalein Island and Roi-Namur, where the wave occurred, are leased by the United States military to serve as the Ronald Reagan Ballistic Missile Defense Test Site .

A view, via aircraft, of the damaged area on the north shore of Roi-Namur, an island in Kwajalein Atoll.

A view, via aircraft, of the damaged area on the north shore of Roi-Namur, an island in Kwajalein Atoll.

Colonel Andrew "Drew" Morgan, Garrison Commander, U.S. Army Garrison - Kwajalein Atoll, said in a Facebook video , "We had a series of unpredicted, gigantic waves wash over the north point of Roi-Namur. It caused extensive damage that we are still assessing, but luckily, there were only a few minor injuries." Another Facebook video showed the coastal damage from a plane .

More than 80 people were evacuated from Roi-Namur back to Kwajalein Island, Morgan said, where more medical services are available. At the site where the wave hit, 50 soldiers were dropped off to fix utilities and assess damages.

"This is going to go down in Kwaj's history books as one of its most challenging times ever in its 80-year history," Morgan said, noting recovery could take months or even years. "We will get through this. Remember, we're important to the nation's security. This is our home. We love our home."

The Marshall Islands, where the wave occurred, are in the Pacific Ocean, northeast of Papua New Guinea and Australia.

The Marshall Islands, where the wave occurred, are in the Pacific Ocean, northeast of Papua New Guinea and Australia.

AccuWeather spoke with National Weather Service (NWS) meteorologist Brandon Aydlett from the NWS Guam office about weather and wave conditions ahead of the incident.

High surf conditions were forecast ahead of the wave

Aydlett said they expected a notable wave event several days in advance due to a massive north Pacific cyclone passing well north of the area. The wind field was unusually large and was followed by an abnormally large area of high pressure. The NWS office in Guam issued high seas and high surf warnings for waves over 15 feet for the islands, which is out of the ordinary.

A NOAA weather map shows the low pressure system that caused the high surf, northeast of the Marshall Islands, located at the &quot;x.&quot;

A weather map shows the low pressure system that caused the high surf, northeast of the Marshall Islands, located at the "x" on Jan. 19, 2024 (NOAA)

What caused the big wave?

Although the NWS office in Guam has received reports of significant inundation from many islands during this event, none have been similar to what was experienced at Roi Namur and shown in the viral video.

While a weather pattern with a strong cyclone and strong high-pressure system was the overall cause, the NWS forecaster's best assessment of what caused the big wave was a complicated story of wave and wind mechanics.

"The motion of the cyclone which, paired with the typhoon-force north winds, likely resulted in a dynamic fetch event," Aydlett told AccuWeather in an email. "This led to a constructive interference of swells of various periods culminating in what occurred at Roi-Namur," Aydlett concluded.

Fetch refers to the distance the wind blows over water in a single direction .

Screen captures from the video of the wave breaking through the doors of a building on Roi-Namur on Jan. 20, 2024 (Erik Hanson)

Screen captures from the video of the wave breaking through the doors of a building on Roi-Namur on Jan. 20, 2024 (Erik Hanson)

What is this phenomenon called?

Although most media outlets are referring to this event as a "rogue wave," rogue waves are large waves that hit ships at sea, not on shore, according to NOAA . This event was most likely a "sneaker wave," defined as an unusually high wave in a set of waves that hit the shore.

Larry Smith, a meteorologist at the NWS office in Monterey, California, said in 2013 , "Though the terms 'sneaker' and 'rogue' wave are often used interchangeably in media reports, Smith considers a 'rogue wave' a different phenomenon, one that occurs out at sea, as a result of wave interactions."

This event was not caused by the storm surge of a tropical storm nor was it a tsunami, which is a large rise in the water level at the shore, usually caused by earthquakes.

Due to the complex nature of winds, waves and bathymetry, or depth of water, individual sneaker waves cannot be predicted today, but recent research indicates that might not always be true. In many cases, however, there were dangerous waves at the locations before sneaker wave damage. To avoid sneaker waves, the NWS advises beachgoers to heed advisories and observe waves before approaching a beach and to be mindful of the ocean.

brain travelling waves

Want next-level safety, ad-free? Unlock advanced, hyperlocal severe weather alerts when you  subscribe to Premium+  on the  AccuWeather app .   AccuWeather Alerts ™  are prompted by our expert meteorologists who monitor and analyze dangerous weather risks 24/7 to keep you and your family safer.

Top Stories

brain travelling waves

Atmospheric river continues to unleash drenching rain in the West

brain travelling waves

Dr. Joel Myers reflects on cold weather and the Challenger disaster

brain travelling waves

All 50 states have seen snow this season

Accuweather early, accuweather prime.

Winter Weather

Clipper storm to spread a quick burst of snow from early to midweek

4 hours ago

brain travelling waves

Woman was stuck overnight on ski resort gondola in Lake Tahoe

5 hours ago

brain travelling waves

Weather News

Billions of cicadas set to appear in rare ‘double brood emergence

brain travelling waves

Reality check: Northeast faces snow, colder air following January thaw

brain travelling waves

Featured Stories

How record-setting cold contributed to Challenger disaster

brain travelling waves

Damaged rotor blade ends NASA’s Ingenuity helicopter mission on Mars

brain travelling waves

Airline COO sits by door during first Boeing flight after grounding

brain travelling waves

PG&E agrees to $45 million settlement related to Dixie Fire

brain travelling waves

We have updated our Privacy Policy and Cookie Policy .

Get AccuWeather alerts as they happen with our browser notifications.

Notifications Enabled

Thanks! We’ll keep you informed.

ComScore

IMAGES

  1. Brain Waves

    brain travelling waves

  2. How brain waves guide memory formation

    brain travelling waves

  3. Brainwaves

    brain travelling waves

  4. What are Brain Waves?

    brain travelling waves

  5. Brainwave Frequencies Explained

    brain travelling waves

  6. Brainwaves and what kind of "role" do they have in our lives

    brain travelling waves

VIDEO

  1. Travelling into the brain

COMMENTS

  1. Traveling and standing waves in the brain

    6 Altmetric Metrics Studying the natural wanderings of the living brain is extremely challenging. Bolt et al. describe a new framework for considering the brain's intrinsic activity based on...

  2. "Traveling" Brain Waves May Be Critical for Cognition

    Now a new study from a team at Columbia University led by neuroscientist Joshua Jacobs suggests traveling waves are widespread in the human cortex—the seat of higher cognitive functions—and...

  3. Traveling waves in the prefrontal cortex during working memory

    Traveling waves are spatially extended patterns in which aligned peaks of activity move sequentially across the cortical surface. Some traveling waves were planar but most rotated. The prefrontal cortex is important for working memory.

  4. "Traveling" nature of brain waves may help working memory work

    Press Inquiries Caption A stadium wave forms when fans in adjacent sections stand up and then sit back down in sequence around the seating area. This creates a wave that travels through the crowd even though no individual fans leave their seats.

  5. 'Traveling' nature of brain waves may help working memory work

    January 31, 2022 Research Findings 'Traveling' nature of brain waves may help working memory work The act of holding information in mind is accompanied by coordination of rotating brain waves in the prefrontal cortex, a new study finds.

  6. Traveling and standing waves in the brain

    35902650 PMC10170397 10.1038/s41593-022-01119- Research Support, N.I.H., Intramural Z99 MH999999/ImNIH/Intramural NIH HHS/United States ZIA MH002783/ImNIH/Intramural NIH HHS/United States

  7. 'Traveling' Nature of Brain Waves May Help Working Memory Work

    An especially underappreciated aspect of the phenomenon is that waves spatially propagate, or "travel," through brain regions over time.

  8. Global waves synchronize the brain's functional systems with

    Four major predictions follow from this traveling wave model. First, BOLD signal fluctuations throughout the brain should be coherent with arousal fluctuations. Second, regional phase shifts of the BOLD signal, relative to physiological indices of arousal, should be organized according to FC network identity.

  9. Traveling brain waves help detect hard-to-see objects

    They found that patterns of neural signals, called traveling brain waves, exist in the visual system of the awake brain and are organized to allow the brain to perceive objects that are faint or otherwise difficult to see. The findings were published in Nature on October 7, 2020. Top from left: Zac Davis and Terrence Sejnowski.

  10. A theoretical basis for standing and traveling brain waves measured

    A theoretical basis for standing and traveling brain waves measured with human EEG with implications for an integrated consciousness Paul L. Nunez a , Ramesh Srinivasan b Add to Mendeley https://doi.org/10.1016/j.clinph.2006.06.754 Get rights and content Methods Propagation velocity Corticocortical fibers Physiological basis for EEG

  11. (PDF) How can we detect and analyze traveling waves in human brain

    Home Travel How can we detect and analyze traveling waves in human brain oscillations? DOI: Authors: Anup Das Columbia University Erfan Zabeh Columbia University Joshua Jacobs Preprints and...

  12. When Brain Waves Go Traveling

    Now Alexander et al. are back with a new PLoS ONE paper in which they describe traveling waves in human brain activity, as measured with magnetoencephalography (MEG). The authors scanned 20 volunteers during a visual and auditory task. Alexander et al. focussed on "the class of waves which are characterized by a linear trajectory in the ...

  13. DMT alters cortical travelling waves

    Quantifying travelling waves. As demonstrated by both theoretical and experimental evidence (Nunez, 2000; Nunez and Srinivasan, 2014; Nunez and Srinivasan, 2009), in most systems, including the human brain, travelling waves occur in groups (or packets) over some range of spatial wavelengths having multiple spatial and temporal frequencies.Given any configurations of electrodes, only parts of ...

  14. A theoretical basis for standing and traveling brain waves measured

    Speculate on the possible roles of traveling and standing waves of synaptic action in facilitating interactions between brain networks. Various features of the global theory of standing and traveling waves, first introduced by the senior author (Nunez, 1972), have been published over the past 34 years.

  15. Theta and Alpha Oscillations Are Traveling Waves in the Human Neocortex

    Human traveling theta and alpha waves can be modeled by a network of coupled oscillators because the direction of wave propagation correlated with the spatial orientation of local frequency gradients. Our findings suggest that oscillations support brain connectivity by organizing neural processes across space and time.

  16. Interictal discharges in the human brain are travelling waves arising

    Although the traditional interpretation of clinically observed ictal discharges in the electrode recordings has been that the underlying brain regions are actively seizing, studies using microelectrode recordings have challenged this interpretation, demonstrating that ictal discharges may in fact reflect receipt of travelling waves, propagated ...

  17. Can quantum hints in the brain revive a radical consciousness theory?

    With anaesthetics and brain organoids, we are finally testing the idea that quantum effects explain consciousness - and the early results suggest this long-derided idea may have been misconstrued

  18. See what your brain does when you look at art

    See what your brain does when you look at art. Headsets that show the impact of art on human brainwaves are to be toured at museums and galleries around the UK. They are connected to an ...

  19. PDF Traveling waves in the prefrontal cortex during working memory

    occurs. Traveling waves have most often been reported in the lower-frequency bands (<30 Hz). Examples include beta-band (15-30 Hz) traveling waves in motor and visual cortices [14,15] and theta band (3-5 Hz) traveling waves in the hippocampus [16,17]. Traveling waves are of interest because they have a variety of useful properties for ...

  20. Here's what happens to your body during plane turbulence

    However, you only have the seat in front of you or the cabin as a reference point, which means your inner ear becomes the dominant sensory message to the brain during turbulence and disrupts the ...

  21. Sound waves get Alzheimer's drug past brain barrier, small ...

    Sound waves get Alzheimer's drug past brain barrier, small study shows. Story by Mark Johnson • 2w. Health Topics mentioned in this article. Mri. Encephalopathy. Brain tumor.

  22. Treating the Brain With Focused Ultrasound

    Focused ultrasound is an early stage, noninvasive therapy with the potential to treat a range of medical conditions. Like diagnostic ultrasound, it uses sound waves above the range of human hearing. But its purpose is to interact with tissues in the body, rather than just produce images of them. In focused ultrasound, multiple, intersecting ...

  23. This Engineer's Hardware Is Inspired by the Brain

    Traveling stuck with Suri in adulthood. He studied and worked in Dubai, the United States, France, and Belgium over the course of his twenties. ... 35 Years Ago, Researchers Used Brain Waves to ...

  24. Insect-inspired AI: how "natural intelligence" could shape the third wave

    "Having found that that simple model of that part of the brain can do image recognition, the industry has then tried to turn that into a silver bullet that solves all the problems in autonomy ...

  25. Racism linked to health risks, brain changes in minorities : Short Wave

    Experiencing racism may physically change your brain. Scientists know that Black people are at a greater risk for health problems like heart disease, diabetes and Alzheimer's disease than white ...

  26. 'Rogue' or 'sneaker?' What caused the giant wave in the ...

    The big wave happened on Saturday, Jan. 20, 2024, on the island of Roi-Namur, part of the Kwajalein Atoll, in the Marshall Islands. The wave caused significant damage to Dyess Army Field and ...